Avalanche: An end-to-end library for continual learning

V. Lomonaco; L. Pellegrini; A. Cossu; A. Carta; G. Graffieti; T.L. Hayes; M. De Lange; M. Masana; J. Pomponi; G.M. Van De Ven; M. Mundt; Q. She; K. Cooper; J. Forest; E. Belouadah; S. Calderara; G.I. Parisi; F. Cuzzolin; A.S. Tolias; S. Scardapane; L. Antiga; S. Ahmad; A. Popescu; C. Kanan; J. Van De Weijer; T. Tuytelaars; D. Bacciu; D. Maltoni

doi:10.1109/CVPRW53098.2021.00399

Avalanche: An end-to-end library for continual learning

V. Lomonaco
, L. Pellegrini
, A. Cossu
, A. Carta
, G. Graffieti
, T.L. Hayes
, M. De Lange
, M. Masana
, J. Pomponi
, G.M. Van De Ven
, M. Mundt
, Q. She
, K. Cooper
, J. Forest
, E. Belouadah
, S. Calderara
, G.I. Parisi
, F. Cuzzolin
, A.S. Tolias
, S. Scardapane

L. Antiga, S. Ahmad, A. Popescu, C. Kanan, J. Van De Weijer, T. Tuytelaars, D. Bacciu, D. Maltoni

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

Learning continually from non-stationary data streams is a long-standing goal and a challenging problem in machine learning. Recently, we have witnessed a renewed and fast-growing interest in continual learning, especially within the deep learning community. However, algorithmic solutions are often difficult to re-implement, evaluate and port across different settings, where even results on standard benchmarks are hard to reproduce. In this work, we propose Avalanche, an open-source end-to-end library for continual learning research based on PyTorch. Avalanche is designed to provide a shared and collaborative codebase for fast prototyping, training, and reproducible evaluation of continual learning algorithms.

Original language	English
Title of host publication	Proceedings - 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2021
Pages	3595-3605
Number of pages	11
ISBN (Electronic)	9781665448994
DOIs	https://doi.org/10.1109/CVPRW53098.2021.00399
Publication status	Published - 2021
Externally published	Yes
Event	2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops: CVPRW 2021 - Virtuell, United States Duration: 19 Jun 2021 → 25 Jun 2021

Conference

Conference	2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops
Abbreviated title	CVPRW 2021
Country/Territory	United States
City	Virtuell
Period	19/06/21 → 25/06/21

ASJC Scopus subject areas

Electrical and Electronic Engineering
Computer Vision and Pattern Recognition

Access to Document

10.1109/CVPRW53098.2021.00399

Cite this

Lomonaco, V., Pellegrini, L., Cossu, A., Carta, A., Graffieti, G., Hayes, T. L., De Lange, M., Masana, M., Pomponi, J., Van De Ven, G. M., Mundt, M., She, Q., Cooper, K., Forest, J., Belouadah, E., Calderara, S., Parisi, G. I., Cuzzolin, F., Tolias, A. S., ... Maltoni, D. (2021). Avalanche: An end-to-end library for continual learning. In Proceedings - 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2021 (pp. 3595-3605) https://doi.org/10.1109/CVPRW53098.2021.00399

Lomonaco, V, Pellegrini, L, Cossu, A, Carta, A, Graffieti, G, Hayes, TL, De Lange, M, Masana, M, Pomponi, J, Van De Ven, GM, Mundt, M, She, Q, Cooper, K, Forest, J, Belouadah, E, Calderara, S, Parisi, GI, Cuzzolin, F, Tolias, AS, Scardapane, S, Antiga, L, Ahmad, S, Popescu, A, Kanan, C, Van De Weijer, J, Tuytelaars, T, Bacciu, D & Maltoni, D 2021, Avalanche: An end-to-end library for continual learning. in Proceedings - 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2021. pp. 3595-3605, 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, Virtuell, United States, 19/06/21. https://doi.org/10.1109/CVPRW53098.2021.00399

@inproceedings{f939f3cea0b34a5dbf603160e1a06936,

title = "Avalanche: An end-to-end library for continual learning",

abstract = "Learning continually from non-stationary data streams is a long-standing goal and a challenging problem in machine learning. Recently, we have witnessed a renewed and fast-growing interest in continual learning, especially within the deep learning community. However, algorithmic solutions are often difficult to re-implement, evaluate and port across different settings, where even results on standard benchmarks are hard to reproduce. In this work, we propose Avalanche, an open-source end-to-end library for continual learning research based on PyTorch. Avalanche is designed to provide a shared and collaborative codebase for fast prototyping, training, and reproducible evaluation of continual learning algorithms.",

author = "V. Lomonaco and L. Pellegrini and A. Cossu and A. Carta and G. Graffieti and T.L. Hayes and \{De Lange\}, M. and M. Masana and J. Pomponi and \{Van De Ven\}, G.M. and M. Mundt and Q. She and K. Cooper and J. Forest and E. Belouadah and S. Calderara and G.I. Parisi and F. Cuzzolin and A.S. Tolias and S. Scardapane and L. Antiga and S. Ahmad and A. Popescu and C. Kanan and \{Van De Weijer\}, J. and T. Tuytelaars and D. Bacciu and D. Maltoni",

year = "2021",

doi = "10.1109/CVPRW53098.2021.00399",

language = "English",

pages = "3595--3605",

booktitle = "Proceedings - 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2021",

note = "2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops : CVPRW 2021, CVPRW 2021 ; Conference date: 19-06-2021 Through 25-06-2021",

}

TY - GEN

T1 - Avalanche: An end-to-end library for continual learning

AU - Lomonaco, V.

AU - Pellegrini, L.

AU - Cossu, A.

AU - Carta, A.

AU - Graffieti, G.

AU - Hayes, T.L.

AU - De Lange, M.

AU - Masana, M.

AU - Pomponi, J.

AU - Van De Ven, G.M.

AU - Mundt, M.

AU - She, Q.

AU - Cooper, K.

AU - Forest, J.

AU - Belouadah, E.

AU - Calderara, S.

AU - Parisi, G.I.

AU - Cuzzolin, F.

AU - Tolias, A.S.

AU - Scardapane, S.

AU - Antiga, L.

AU - Ahmad, S.

AU - Popescu, A.

AU - Kanan, C.

AU - Van De Weijer, J.

AU - Tuytelaars, T.

AU - Bacciu, D.

AU - Maltoni, D.

PY - 2021

Y1 - 2021

N2 - Learning continually from non-stationary data streams is a long-standing goal and a challenging problem in machine learning. Recently, we have witnessed a renewed and fast-growing interest in continual learning, especially within the deep learning community. However, algorithmic solutions are often difficult to re-implement, evaluate and port across different settings, where even results on standard benchmarks are hard to reproduce. In this work, we propose Avalanche, an open-source end-to-end library for continual learning research based on PyTorch. Avalanche is designed to provide a shared and collaborative codebase for fast prototyping, training, and reproducible evaluation of continual learning algorithms.

AB - Learning continually from non-stationary data streams is a long-standing goal and a challenging problem in machine learning. Recently, we have witnessed a renewed and fast-growing interest in continual learning, especially within the deep learning community. However, algorithmic solutions are often difficult to re-implement, evaluate and port across different settings, where even results on standard benchmarks are hard to reproduce. In this work, we propose Avalanche, an open-source end-to-end library for continual learning research based on PyTorch. Avalanche is designed to provide a shared and collaborative codebase for fast prototyping, training, and reproducible evaluation of continual learning algorithms.

UR - https://www.scopus.com/pages/publications/85107928433

U2 - 10.1109/CVPRW53098.2021.00399

DO - 10.1109/CVPRW53098.2021.00399

M3 - Conference paper

SP - 3595

EP - 3605

BT - Proceedings - 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2021

T2 - 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops

Y2 - 19 June 2021 through 25 June 2021

ER -

Avalanche: An end-to-end library for continual learning

Abstract

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Cite this