Efficient Perturbation Inference and Expandable Network for continual learning

Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and...
Ausführliche Beschreibung

Gespeichert in:

Autor*in:	Du, Fei [verfasserIn] Yang, Yun [verfasserIn] Zhao, Ziyuan [verfasserIn] Zeng, Zeng [verfasserIn]

Format:	E-Artikel
Sprache:	Englisch

Erschienen:	2022

Schlagwörter:	Continual learning Dynamic networks Class incremental learning Uncertainty inference

Übergeordnetes Werk:	Enthalten in: Neural networks - Amsterdam : Elsevier, 1988, 159, Seite 97-106
Übergeordnetes Werk:	volume:159 ; pages:97-106

DOI / URN:	10.1016/j.neunet.2022.10.030

Katalog-ID:	ELV009155317

Internformat


LEADER	01000caa a22002652 4500
001	ELV009155317
003	DE-627
005	20230524162716.0
007	cr uuu---uuuuu
008	230510s2022 xx \|\|\|\|\|o 00\| \|\|eng c
024	7		\|a 10.1016/j.neunet.2022.10.030 \|2 doi
035			\|a (DE-627)ELV009155317
035			\|a (ELSEVIER)S0893-6080(22)00426-9
040			\|a DE-627 \|b ger \|c DE-627 \|e rda
041			\|a eng
082	0	4	\|a 004 \|q DE-600
084			\|a 54.72 \|2 bkl
100	1		\|a Du, Fei \|e verfasserin \|4 aut
245	1	0	\|a Efficient Perturbation Inference and Expandable Network for continual learning
264		1	\|c 2022
336			\|a nicht spezifiziert \|b zzz \|2 rdacontent
337			\|a Computermedien \|b c \|2 rdamedia
338			\|a Online-Ressource \|b cr \|2 rdacarrier
520			\|a Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters.
650		4	\|a Continual learning
650		4	\|a Dynamic networks
650		4	\|a Class incremental learning
650		4	\|a Uncertainty inference
700	1		\|a Yang, Yun \|e verfasserin \|0 (orcid)0000-0002-9893-3436 \|4 aut
700	1		\|a Zhao, Ziyuan \|e verfasserin \|4 aut
700	1		\|a Zeng, Zeng \|e verfasserin \|4 aut
773	0	8	\|i Enthalten in \|t Neural networks \|d Amsterdam : Elsevier, 1988 \|g 159, Seite 97-106 \|h Online-Ressource \|w (DE-627)302468536 \|w (DE-600)1491372-0 \|w (DE-576)07971997X \|x 1879-2782 \|7 nnns
773	1	8	\|g volume:159 \|g pages:97-106
912			\|a GBV_USEFLAG_U
912			\|a SYSFLAG_U
912			\|a GBV_ELV
912			\|a GBV_ILN_20
912			\|a GBV_ILN_22
912			\|a GBV_ILN_23
912			\|a GBV_ILN_24
912			\|a GBV_ILN_31
912			\|a GBV_ILN_32
912			\|a GBV_ILN_40
912			\|a GBV_ILN_60
912			\|a GBV_ILN_62
912			\|a GBV_ILN_63
912			\|a GBV_ILN_65
912			\|a GBV_ILN_69
912			\|a GBV_ILN_70
912			\|a GBV_ILN_73
912			\|a GBV_ILN_74
912			\|a GBV_ILN_90
912			\|a GBV_ILN_95
912			\|a GBV_ILN_100
912			\|a GBV_ILN_101
912			\|a GBV_ILN_105
912			\|a GBV_ILN_110
912			\|a GBV_ILN_150
912			\|a GBV_ILN_151
912			\|a GBV_ILN_224
912			\|a GBV_ILN_370
912			\|a GBV_ILN_602
912			\|a GBV_ILN_702
912			\|a GBV_ILN_2003
912			\|a GBV_ILN_2004
912			\|a GBV_ILN_2005
912			\|a GBV_ILN_2011
912			\|a GBV_ILN_2014
912			\|a GBV_ILN_2015
912			\|a GBV_ILN_2020
912			\|a GBV_ILN_2021
912			\|a GBV_ILN_2025
912			\|a GBV_ILN_2027
912			\|a GBV_ILN_2034
912			\|a GBV_ILN_2038
912			\|a GBV_ILN_2044
912			\|a GBV_ILN_2048
912			\|a GBV_ILN_2049
912			\|a GBV_ILN_2050
912			\|a GBV_ILN_2056
912			\|a GBV_ILN_2059
912			\|a GBV_ILN_2061
912			\|a GBV_ILN_2064
912			\|a GBV_ILN_2065
912			\|a GBV_ILN_2068
912			\|a GBV_ILN_2111
912			\|a GBV_ILN_2112
912			\|a GBV_ILN_2113
912			\|a GBV_ILN_2118
912			\|a GBV_ILN_2122
912			\|a GBV_ILN_2129
912			\|a GBV_ILN_2143
912			\|a GBV_ILN_2147
912			\|a GBV_ILN_2148
912			\|a GBV_ILN_2152
912			\|a GBV_ILN_2153
912			\|a GBV_ILN_2190
912			\|a GBV_ILN_2336
912			\|a GBV_ILN_2507
912			\|a GBV_ILN_2522
912			\|a GBV_ILN_4035
912			\|a GBV_ILN_4037
912			\|a GBV_ILN_4112
912			\|a GBV_ILN_4125
912			\|a GBV_ILN_4126
912			\|a GBV_ILN_4242
912			\|a GBV_ILN_4251
912			\|a GBV_ILN_4305
912			\|a GBV_ILN_4313
912			\|a GBV_ILN_4323
912			\|a GBV_ILN_4324
912			\|a GBV_ILN_4326
912			\|a GBV_ILN_4333
912			\|a GBV_ILN_4334
912			\|a GBV_ILN_4335
912			\|a GBV_ILN_4338
912			\|a GBV_ILN_4393
936	b	k	\|a 54.72 \|j Künstliche Intelligenz
951			\|a AR
952			\|d 159 \|h 97-106

Indexfelder

author_variant	f d fd y y yy z z zz z z zz
matchkey_str	article:18792782:2022----::fiinprubtoifrnenepnalntok
hierarchy_sort_str	2022
bklnumber	54.72
publishDate	2022
allfields	10.1016/j.neunet.2022.10.030 doi (DE-627)ELV009155317 (ELSEVIER)S0893-6080(22)00426-9 DE-627 ger DE-627 rda eng 004 DE-600 54.72 bkl Du, Fei verfasserin aut Efficient Perturbation Inference and Expandable Network for continual learning 2022 nicht spezifiziert zzz rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters. Continual learning Dynamic networks Class incremental learning Uncertainty inference Yang, Yun verfasserin (orcid)0000-0002-9893-3436 aut Zhao, Ziyuan verfasserin aut Zeng, Zeng verfasserin aut Enthalten in Neural networks Amsterdam : Elsevier, 1988 159, Seite 97-106 Online-Ressource (DE-627)302468536 (DE-600)1491372-0 (DE-576)07971997X 1879-2782 nnns volume:159 pages:97-106 GBV_USEFLAG_U SYSFLAG_U GBV_ELV GBV_ILN_20 GBV_ILN_22 GBV_ILN_23 GBV_ILN_24 GBV_ILN_31 GBV_ILN_32 GBV_ILN_40 GBV_ILN_60 GBV_ILN_62 GBV_ILN_63 GBV_ILN_65 GBV_ILN_69 GBV_ILN_70 GBV_ILN_73 GBV_ILN_74 GBV_ILN_90 GBV_ILN_95 GBV_ILN_100 GBV_ILN_101 GBV_ILN_105 GBV_ILN_110 GBV_ILN_150 GBV_ILN_151 GBV_ILN_224 GBV_ILN_370 GBV_ILN_602 GBV_ILN_702 GBV_ILN_2003 GBV_ILN_2004 GBV_ILN_2005 GBV_ILN_2011 GBV_ILN_2014 GBV_ILN_2015 GBV_ILN_2020 GBV_ILN_2021 GBV_ILN_2025 GBV_ILN_2027 GBV_ILN_2034 GBV_ILN_2038 GBV_ILN_2044 GBV_ILN_2048 GBV_ILN_2049 GBV_ILN_2050 GBV_ILN_2056 GBV_ILN_2059 GBV_ILN_2061 GBV_ILN_2064 GBV_ILN_2065 GBV_ILN_2068 GBV_ILN_2111 GBV_ILN_2112 GBV_ILN_2113 GBV_ILN_2118 GBV_ILN_2122 GBV_ILN_2129 GBV_ILN_2143 GBV_ILN_2147 GBV_ILN_2148 GBV_ILN_2152 GBV_ILN_2153 GBV_ILN_2190 GBV_ILN_2336 GBV_ILN_2507 GBV_ILN_2522 GBV_ILN_4035 GBV_ILN_4037 GBV_ILN_4112 GBV_ILN_4125 GBV_ILN_4126 GBV_ILN_4242 GBV_ILN_4251 GBV_ILN_4305 GBV_ILN_4313 GBV_ILN_4323 GBV_ILN_4324 GBV_ILN_4326 GBV_ILN_4333 GBV_ILN_4334 GBV_ILN_4335 GBV_ILN_4338 GBV_ILN_4393 54.72 Künstliche Intelligenz AR 159 97-106
spelling	10.1016/j.neunet.2022.10.030 doi (DE-627)ELV009155317 (ELSEVIER)S0893-6080(22)00426-9 DE-627 ger DE-627 rda eng 004 DE-600 54.72 bkl Du, Fei verfasserin aut Efficient Perturbation Inference and Expandable Network for continual learning 2022 nicht spezifiziert zzz rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters. Continual learning Dynamic networks Class incremental learning Uncertainty inference Yang, Yun verfasserin (orcid)0000-0002-9893-3436 aut Zhao, Ziyuan verfasserin aut Zeng, Zeng verfasserin aut Enthalten in Neural networks Amsterdam : Elsevier, 1988 159, Seite 97-106 Online-Ressource (DE-627)302468536 (DE-600)1491372-0 (DE-576)07971997X 1879-2782 nnns volume:159 pages:97-106 GBV_USEFLAG_U SYSFLAG_U GBV_ELV GBV_ILN_20 GBV_ILN_22 GBV_ILN_23 GBV_ILN_24 GBV_ILN_31 GBV_ILN_32 GBV_ILN_40 GBV_ILN_60 GBV_ILN_62 GBV_ILN_63 GBV_ILN_65 GBV_ILN_69 GBV_ILN_70 GBV_ILN_73 GBV_ILN_74 GBV_ILN_90 GBV_ILN_95 GBV_ILN_100 GBV_ILN_101 GBV_ILN_105 GBV_ILN_110 GBV_ILN_150 GBV_ILN_151 GBV_ILN_224 GBV_ILN_370 GBV_ILN_602 GBV_ILN_702 GBV_ILN_2003 GBV_ILN_2004 GBV_ILN_2005 GBV_ILN_2011 GBV_ILN_2014 GBV_ILN_2015 GBV_ILN_2020 GBV_ILN_2021 GBV_ILN_2025 GBV_ILN_2027 GBV_ILN_2034 GBV_ILN_2038 GBV_ILN_2044 GBV_ILN_2048 GBV_ILN_2049 GBV_ILN_2050 GBV_ILN_2056 GBV_ILN_2059 GBV_ILN_2061 GBV_ILN_2064 GBV_ILN_2065 GBV_ILN_2068 GBV_ILN_2111 GBV_ILN_2112 GBV_ILN_2113 GBV_ILN_2118 GBV_ILN_2122 GBV_ILN_2129 GBV_ILN_2143 GBV_ILN_2147 GBV_ILN_2148 GBV_ILN_2152 GBV_ILN_2153 GBV_ILN_2190 GBV_ILN_2336 GBV_ILN_2507 GBV_ILN_2522 GBV_ILN_4035 GBV_ILN_4037 GBV_ILN_4112 GBV_ILN_4125 GBV_ILN_4126 GBV_ILN_4242 GBV_ILN_4251 GBV_ILN_4305 GBV_ILN_4313 GBV_ILN_4323 GBV_ILN_4324 GBV_ILN_4326 GBV_ILN_4333 GBV_ILN_4334 GBV_ILN_4335 GBV_ILN_4338 GBV_ILN_4393 54.72 Künstliche Intelligenz AR 159 97-106
allfields_unstemmed	10.1016/j.neunet.2022.10.030 doi (DE-627)ELV009155317 (ELSEVIER)S0893-6080(22)00426-9 DE-627 ger DE-627 rda eng 004 DE-600 54.72 bkl Du, Fei verfasserin aut Efficient Perturbation Inference and Expandable Network for continual learning 2022 nicht spezifiziert zzz rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters. Continual learning Dynamic networks Class incremental learning Uncertainty inference Yang, Yun verfasserin (orcid)0000-0002-9893-3436 aut Zhao, Ziyuan verfasserin aut Zeng, Zeng verfasserin aut Enthalten in Neural networks Amsterdam : Elsevier, 1988 159, Seite 97-106 Online-Ressource (DE-627)302468536 (DE-600)1491372-0 (DE-576)07971997X 1879-2782 nnns volume:159 pages:97-106 GBV_USEFLAG_U SYSFLAG_U GBV_ELV GBV_ILN_20 GBV_ILN_22 GBV_ILN_23 GBV_ILN_24 GBV_ILN_31 GBV_ILN_32 GBV_ILN_40 GBV_ILN_60 GBV_ILN_62 GBV_ILN_63 GBV_ILN_65 GBV_ILN_69 GBV_ILN_70 GBV_ILN_73 GBV_ILN_74 GBV_ILN_90 GBV_ILN_95 GBV_ILN_100 GBV_ILN_101 GBV_ILN_105 GBV_ILN_110 GBV_ILN_150 GBV_ILN_151 GBV_ILN_224 GBV_ILN_370 GBV_ILN_602 GBV_ILN_702 GBV_ILN_2003 GBV_ILN_2004 GBV_ILN_2005 GBV_ILN_2011 GBV_ILN_2014 GBV_ILN_2015 GBV_ILN_2020 GBV_ILN_2021 GBV_ILN_2025 GBV_ILN_2027 GBV_ILN_2034 GBV_ILN_2038 GBV_ILN_2044 GBV_ILN_2048 GBV_ILN_2049 GBV_ILN_2050 GBV_ILN_2056 GBV_ILN_2059 GBV_ILN_2061 GBV_ILN_2064 GBV_ILN_2065 GBV_ILN_2068 GBV_ILN_2111 GBV_ILN_2112 GBV_ILN_2113 GBV_ILN_2118 GBV_ILN_2122 GBV_ILN_2129 GBV_ILN_2143 GBV_ILN_2147 GBV_ILN_2148 GBV_ILN_2152 GBV_ILN_2153 GBV_ILN_2190 GBV_ILN_2336 GBV_ILN_2507 GBV_ILN_2522 GBV_ILN_4035 GBV_ILN_4037 GBV_ILN_4112 GBV_ILN_4125 GBV_ILN_4126 GBV_ILN_4242 GBV_ILN_4251 GBV_ILN_4305 GBV_ILN_4313 GBV_ILN_4323 GBV_ILN_4324 GBV_ILN_4326 GBV_ILN_4333 GBV_ILN_4334 GBV_ILN_4335 GBV_ILN_4338 GBV_ILN_4393 54.72 Künstliche Intelligenz AR 159 97-106
allfieldsGer	10.1016/j.neunet.2022.10.030 doi (DE-627)ELV009155317 (ELSEVIER)S0893-6080(22)00426-9 DE-627 ger DE-627 rda eng 004 DE-600 54.72 bkl Du, Fei verfasserin aut Efficient Perturbation Inference and Expandable Network for continual learning 2022 nicht spezifiziert zzz rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters. Continual learning Dynamic networks Class incremental learning Uncertainty inference Yang, Yun verfasserin (orcid)0000-0002-9893-3436 aut Zhao, Ziyuan verfasserin aut Zeng, Zeng verfasserin aut Enthalten in Neural networks Amsterdam : Elsevier, 1988 159, Seite 97-106 Online-Ressource (DE-627)302468536 (DE-600)1491372-0 (DE-576)07971997X 1879-2782 nnns volume:159 pages:97-106 GBV_USEFLAG_U SYSFLAG_U GBV_ELV GBV_ILN_20 GBV_ILN_22 GBV_ILN_23 GBV_ILN_24 GBV_ILN_31 GBV_ILN_32 GBV_ILN_40 GBV_ILN_60 GBV_ILN_62 GBV_ILN_63 GBV_ILN_65 GBV_ILN_69 GBV_ILN_70 GBV_ILN_73 GBV_ILN_74 GBV_ILN_90 GBV_ILN_95 GBV_ILN_100 GBV_ILN_101 GBV_ILN_105 GBV_ILN_110 GBV_ILN_150 GBV_ILN_151 GBV_ILN_224 GBV_ILN_370 GBV_ILN_602 GBV_ILN_702 GBV_ILN_2003 GBV_ILN_2004 GBV_ILN_2005 GBV_ILN_2011 GBV_ILN_2014 GBV_ILN_2015 GBV_ILN_2020 GBV_ILN_2021 GBV_ILN_2025 GBV_ILN_2027 GBV_ILN_2034 GBV_ILN_2038 GBV_ILN_2044 GBV_ILN_2048 GBV_ILN_2049 GBV_ILN_2050 GBV_ILN_2056 GBV_ILN_2059 GBV_ILN_2061 GBV_ILN_2064 GBV_ILN_2065 GBV_ILN_2068 GBV_ILN_2111 GBV_ILN_2112 GBV_ILN_2113 GBV_ILN_2118 GBV_ILN_2122 GBV_ILN_2129 GBV_ILN_2143 GBV_ILN_2147 GBV_ILN_2148 GBV_ILN_2152 GBV_ILN_2153 GBV_ILN_2190 GBV_ILN_2336 GBV_ILN_2507 GBV_ILN_2522 GBV_ILN_4035 GBV_ILN_4037 GBV_ILN_4112 GBV_ILN_4125 GBV_ILN_4126 GBV_ILN_4242 GBV_ILN_4251 GBV_ILN_4305 GBV_ILN_4313 GBV_ILN_4323 GBV_ILN_4324 GBV_ILN_4326 GBV_ILN_4333 GBV_ILN_4334 GBV_ILN_4335 GBV_ILN_4338 GBV_ILN_4393 54.72 Künstliche Intelligenz AR 159 97-106
allfieldsSound	10.1016/j.neunet.2022.10.030 doi (DE-627)ELV009155317 (ELSEVIER)S0893-6080(22)00426-9 DE-627 ger DE-627 rda eng 004 DE-600 54.72 bkl Du, Fei verfasserin aut Efficient Perturbation Inference and Expandable Network for continual learning 2022 nicht spezifiziert zzz rdacontent Computermedien c rdamedia Online-Ressource cr rdacarrier Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters. Continual learning Dynamic networks Class incremental learning Uncertainty inference Yang, Yun verfasserin (orcid)0000-0002-9893-3436 aut Zhao, Ziyuan verfasserin aut Zeng, Zeng verfasserin aut Enthalten in Neural networks Amsterdam : Elsevier, 1988 159, Seite 97-106 Online-Ressource (DE-627)302468536 (DE-600)1491372-0 (DE-576)07971997X 1879-2782 nnns volume:159 pages:97-106 GBV_USEFLAG_U SYSFLAG_U GBV_ELV GBV_ILN_20 GBV_ILN_22 GBV_ILN_23 GBV_ILN_24 GBV_ILN_31 GBV_ILN_32 GBV_ILN_40 GBV_ILN_60 GBV_ILN_62 GBV_ILN_63 GBV_ILN_65 GBV_ILN_69 GBV_ILN_70 GBV_ILN_73 GBV_ILN_74 GBV_ILN_90 GBV_ILN_95 GBV_ILN_100 GBV_ILN_101 GBV_ILN_105 GBV_ILN_110 GBV_ILN_150 GBV_ILN_151 GBV_ILN_224 GBV_ILN_370 GBV_ILN_602 GBV_ILN_702 GBV_ILN_2003 GBV_ILN_2004 GBV_ILN_2005 GBV_ILN_2011 GBV_ILN_2014 GBV_ILN_2015 GBV_ILN_2020 GBV_ILN_2021 GBV_ILN_2025 GBV_ILN_2027 GBV_ILN_2034 GBV_ILN_2038 GBV_ILN_2044 GBV_ILN_2048 GBV_ILN_2049 GBV_ILN_2050 GBV_ILN_2056 GBV_ILN_2059 GBV_ILN_2061 GBV_ILN_2064 GBV_ILN_2065 GBV_ILN_2068 GBV_ILN_2111 GBV_ILN_2112 GBV_ILN_2113 GBV_ILN_2118 GBV_ILN_2122 GBV_ILN_2129 GBV_ILN_2143 GBV_ILN_2147 GBV_ILN_2148 GBV_ILN_2152 GBV_ILN_2153 GBV_ILN_2190 GBV_ILN_2336 GBV_ILN_2507 GBV_ILN_2522 GBV_ILN_4035 GBV_ILN_4037 GBV_ILN_4112 GBV_ILN_4125 GBV_ILN_4126 GBV_ILN_4242 GBV_ILN_4251 GBV_ILN_4305 GBV_ILN_4313 GBV_ILN_4323 GBV_ILN_4324 GBV_ILN_4326 GBV_ILN_4333 GBV_ILN_4334 GBV_ILN_4335 GBV_ILN_4338 GBV_ILN_4393 54.72 Künstliche Intelligenz AR 159 97-106
language	English
source	Enthalten in Neural networks 159, Seite 97-106 volume:159 pages:97-106
sourceStr	Enthalten in Neural networks 159, Seite 97-106 volume:159 pages:97-106
format_phy_str_mv	Article
bklname	Künstliche Intelligenz
institution	findex.gbv.de
topic_facet	Continual learning Dynamic networks Class incremental learning Uncertainty inference
dewey-raw	004
isfreeaccess_bool	false
container_title	Neural networks
authorswithroles_txt_mv	Du, Fei @@aut@@ Yang, Yun @@aut@@ Zhao, Ziyuan @@aut@@ Zeng, Zeng @@aut@@
publishDateDaySort_date	2022-01-01T00:00:00Z
hierarchy_top_id	302468536
dewey-sort	14
id	ELV009155317
language_de	englisch
fullrecord	<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">ELV009155317</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230524162716.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">230510s2022 xx \|\|\|\|\|o 00\| \|\|eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1016/j.neunet.2022.10.030</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)ELV009155317</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ELSEVIER)S0893-6080(22)00426-9</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rda</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">004</subfield><subfield code="q">DE-600</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.72</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Du, Fei</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Efficient Perturbation Inference and Expandable Network for continual learning</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zzz</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Continual learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Dynamic networks</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Class incremental learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Uncertainty inference</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Yang, Yun</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0002-9893-3436</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhao, Ziyuan</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zeng, Zeng</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Neural networks</subfield><subfield code="d">Amsterdam : Elsevier, 1988</subfield><subfield code="g">159, Seite 97-106</subfield><subfield code="h">Online-Ressource</subfield><subfield code="w">(DE-627)302468536</subfield><subfield code="w">(DE-600)1491372-0</subfield><subfield code="w">(DE-576)07971997X</subfield><subfield code="x">1879-2782</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:159</subfield><subfield code="g">pages:97-106</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ELV</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_20</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_22</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_23</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_24</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_31</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_32</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_40</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_60</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_62</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_63</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_65</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_69</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_73</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_74</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_90</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_95</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_100</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_101</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_105</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_110</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_150</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_151</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_224</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_370</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_602</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_702</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2003</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2004</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2005</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2011</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2014</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2015</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2020</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2021</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2025</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2027</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2034</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2038</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2044</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2048</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2049</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2050</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2056</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2059</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2061</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2064</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2065</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2068</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2111</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2112</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2113</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2118</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2122</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2129</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2143</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2147</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2148</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2152</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2153</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2190</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2336</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2507</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2522</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4035</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4037</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4112</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4125</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4126</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4242</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4251</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4305</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4313</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4323</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4324</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4326</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4333</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4334</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4335</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4338</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4393</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.72</subfield><subfield code="j">Künstliche Intelligenz</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">159</subfield><subfield code="h">97-106</subfield></datafield></record></collection>
author	Du, Fei
spellingShingle	Du, Fei ddc 004 bkl 54.72 misc Continual learning misc Dynamic networks misc Class incremental learning misc Uncertainty inference Efficient Perturbation Inference and Expandable Network for continual learning
authorStr	Du, Fei
ppnlink_with_tag_str_mv	@@773@@(DE-627)302468536
format	electronic Article
dewey-ones	004 - Data processing & computer science
delete_txt_mv	keep
author_role	aut aut aut aut
collection	elsevier
remote_str	true
illustrated	Not Illustrated
issn	1879-2782
topic_title	004 DE-600 54.72 bkl Efficient Perturbation Inference and Expandable Network for continual learning Continual learning Dynamic networks Class incremental learning Uncertainty inference
topic	ddc 004 bkl 54.72 misc Continual learning misc Dynamic networks misc Class incremental learning misc Uncertainty inference
topic_unstemmed	ddc 004 bkl 54.72 misc Continual learning misc Dynamic networks misc Class incremental learning misc Uncertainty inference
topic_browse	ddc 004 bkl 54.72 misc Continual learning misc Dynamic networks misc Class incremental learning misc Uncertainty inference
format_facet	Elektronische Aufsätze Aufsätze Elektronische Ressource
format_main_str_mv	Text Zeitschrift/Artikel
carriertype_str_mv	cr
hierarchy_parent_title	Neural networks
hierarchy_parent_id	302468536
dewey-tens	000 - Computer science, knowledge & systems
hierarchy_top_title	Neural networks
isfreeaccess_txt	false
familylinks_str_mv	(DE-627)302468536 (DE-600)1491372-0 (DE-576)07971997X
title	Efficient Perturbation Inference and Expandable Network for continual learning
ctrlnum	(DE-627)ELV009155317 (ELSEVIER)S0893-6080(22)00426-9
title_full	Efficient Perturbation Inference and Expandable Network for continual learning
author_sort	Du, Fei
journal	Neural networks
journalStr	Neural networks
lang_code	eng
isOA_bool	false
dewey-hundreds	000 - Computer science, information & general works
recordtype	marc
publishDateSort	2022
contenttype_str_mv	zzz
container_start_page	97
author_browse	Du, Fei Yang, Yun Zhao, Ziyuan Zeng, Zeng
container_volume	159
class	004 DE-600 54.72 bkl
format_se	Elektronische Aufsätze
author-letter	Du, Fei
doi_str_mv	10.1016/j.neunet.2022.10.030
normlink	(ORCID)0000-0002-9893-3436
normlink_prefix_str_mv	(orcid)0000-0002-9893-3436
dewey-full	004
author2-role	verfasserin
title_sort	efficient perturbation inference and expandable network for continual learning
title_auth	Efficient Perturbation Inference and Expandable Network for continual learning
abstract	Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters.
abstractGer	Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters.
abstract_unstemmed	Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters.
collection_details	GBV_USEFLAG_U SYSFLAG_U GBV_ELV GBV_ILN_20 GBV_ILN_22 GBV_ILN_23 GBV_ILN_24 GBV_ILN_31 GBV_ILN_32 GBV_ILN_40 GBV_ILN_60 GBV_ILN_62 GBV_ILN_63 GBV_ILN_65 GBV_ILN_69 GBV_ILN_70 GBV_ILN_73 GBV_ILN_74 GBV_ILN_90 GBV_ILN_95 GBV_ILN_100 GBV_ILN_101 GBV_ILN_105 GBV_ILN_110 GBV_ILN_150 GBV_ILN_151 GBV_ILN_224 GBV_ILN_370 GBV_ILN_602 GBV_ILN_702 GBV_ILN_2003 GBV_ILN_2004 GBV_ILN_2005 GBV_ILN_2011 GBV_ILN_2014 GBV_ILN_2015 GBV_ILN_2020 GBV_ILN_2021 GBV_ILN_2025 GBV_ILN_2027 GBV_ILN_2034 GBV_ILN_2038 GBV_ILN_2044 GBV_ILN_2048 GBV_ILN_2049 GBV_ILN_2050 GBV_ILN_2056 GBV_ILN_2059 GBV_ILN_2061 GBV_ILN_2064 GBV_ILN_2065 GBV_ILN_2068 GBV_ILN_2111 GBV_ILN_2112 GBV_ILN_2113 GBV_ILN_2118 GBV_ILN_2122 GBV_ILN_2129 GBV_ILN_2143 GBV_ILN_2147 GBV_ILN_2148 GBV_ILN_2152 GBV_ILN_2153 GBV_ILN_2190 GBV_ILN_2336 GBV_ILN_2507 GBV_ILN_2522 GBV_ILN_4035 GBV_ILN_4037 GBV_ILN_4112 GBV_ILN_4125 GBV_ILN_4126 GBV_ILN_4242 GBV_ILN_4251 GBV_ILN_4305 GBV_ILN_4313 GBV_ILN_4323 GBV_ILN_4324 GBV_ILN_4326 GBV_ILN_4333 GBV_ILN_4334 GBV_ILN_4335 GBV_ILN_4338 GBV_ILN_4393
title_short	Efficient Perturbation Inference and Expandable Network for continual learning
remote_bool	true
author2	Yang, Yun Zhao, Ziyuan Zeng, Zeng
author2Str	Yang, Yun Zhao, Ziyuan Zeng, Zeng
ppnlink	302468536
mediatype_str_mv	c
isOA_txt	false
hochschulschrift_bool	false
doi_str	10.1016/j.neunet.2022.10.030
up_date	2024-07-06T22:11:12.022Z
_version_	1803869352910389248
fullrecord_marcxml	<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">ELV009155317</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230524162716.0</controlfield><controlfield tag="007">cr uuu---uuuuu</controlfield><controlfield tag="008">230510s2022 xx \|\|\|\|\|o 00\| \|\|eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1016/j.neunet.2022.10.030</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)ELV009155317</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(ELSEVIER)S0893-6080(22)00426-9</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rda</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">004</subfield><subfield code="q">DE-600</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.72</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Du, Fei</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Efficient Perturbation Inference and Expandable Network for continual learning</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2022</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">nicht spezifiziert</subfield><subfield code="b">zzz</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">Computermedien</subfield><subfield code="b">c</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Online-Ressource</subfield><subfield code="b">cr</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Although humans are capable of learning new tasks without forgetting previous ones, most neural networks fail to do so because learning new tasks could override the knowledge acquired from previous data. In this work, we alleviate this issue by proposing a novel Efficient Perturbation Inference and Expandable Network (EPIE-Net), which dynamically expands lightweight task-specific decoders for new classes and utilizes a mixed-label uncertainty strategy to improve the robustness. Moreover, we calculate the average probability of perturbed samples at inference, which can generally improve the performance of the model. Experimental results show that our method consistently outperforms other methods with fewer parameters in class incremental learning benchmarks. For example, on the CIFAR-100 10 steps setup, our method achieves an average accuracy of 76.33% and the last accuracy of 65.93% within only 3.46M average parameters.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Continual learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Dynamic networks</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Class incremental learning</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Uncertainty inference</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Yang, Yun</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0002-9893-3436</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhao, Ziyuan</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zeng, Zeng</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Neural networks</subfield><subfield code="d">Amsterdam : Elsevier, 1988</subfield><subfield code="g">159, Seite 97-106</subfield><subfield code="h">Online-Ressource</subfield><subfield code="w">(DE-627)302468536</subfield><subfield code="w">(DE-600)1491372-0</subfield><subfield code="w">(DE-576)07971997X</subfield><subfield code="x">1879-2782</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:159</subfield><subfield code="g">pages:97-106</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_U</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ELV</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_20</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_22</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_23</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_24</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_31</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_32</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_40</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_60</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_62</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_63</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_65</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_69</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_73</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_74</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_90</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_95</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_100</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_101</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_105</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_110</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_150</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_151</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_224</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_370</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_602</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_702</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2003</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2004</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2005</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2011</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2014</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2015</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2020</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2021</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2025</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2027</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2034</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2038</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2044</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2048</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2049</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2050</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2056</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2059</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2061</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2064</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2065</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2068</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2111</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2112</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2113</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2118</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2122</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2129</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2143</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2147</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2148</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2152</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2153</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2190</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2336</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2507</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2522</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4035</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4037</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4112</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4125</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4126</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4242</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4251</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4305</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4313</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4323</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4324</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4326</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4333</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4334</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4335</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4338</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4393</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.72</subfield><subfield code="j">Künstliche Intelligenz</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">159</subfield><subfield code="h">97-106</subfield></datafield></record></collection>
score	7.4006395

Nicht das Richtige dabei?

Schreiben Sie uns!

Efficient Perturbation Inference and Expandable Network for continual learning

Nicht das Richtige dabei?

Zugang & Verfügbarkeit

Vorhandene Bände

Nicht das Richtige dabei?