Active Class Selection for Few-Shot Class-Incremental Learning

Christopher McClurg, Ali Ayub, Harsh Tyagi, Sarah M. Rajtmajer, Alan R. Wagner
Proceedings of The 2nd Conference on Lifelong Learning Agents, PMLR 232:811-827, 2023.

Abstract

For real-world applications, robots will need to continually learn in their environments through limited interactions with their users. Toward this, previous works in few-shot class incremental learning (FSCIL) and active class selection (ACS) have achieved promising results but were tested in constrained setups. Therefore, in this paper, we combine ideas from FSCIL and ACS to develop a novel framework that can allow an autonomous agent to continually learn new objects by asking its users to label only a few of the most informative objects in the environment. To this end, we build on a state-of-the-art (SOTA) FSCIL model and extend it with techniques from ACS literature. We term this model Few-shot Incremental Active class SeleCtiOn (FIASco). We further integrate a potential field-based navigation technique with our model to develop a complete framework that can allow an agent to process and reason on its sensory data through the FIASco model, navigate towards the most informative object in the environment, gather data about the object through its sensors and incrementally update the FIASco model. Experimental results on a simulated agent and a real robot show the significance of our approach for long-term real-world robotics applications.

Cite this Paper

BibTeX
@InProceedings{pmlr-v232-mcclurg23a, title = {Active Class Selection for Few-Shot Class-Incremental Learning}, author = {McClurg, Christopher and Ayub, Ali and Tyagi, Harsh and Rajtmajer, Sarah M. and Wagner, Alan R.}, booktitle = {Proceedings of The 2nd Conference on Lifelong Learning Agents}, pages = {811--827}, year = {2023}, editor = {Chandar, Sarath and Pascanu, Razvan and Sedghi, Hanie and Precup, Doina}, volume = {232}, series = {Proceedings of Machine Learning Research}, month = {22--25 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v232/mcclurg23a/mcclurg23a.pdf}, url = {https://proceedings.mlr.press/v232/mcclurg23a.html}, abstract = {For real-world applications, robots will need to continually learn in their environments through limited interactions with their users. Toward this, previous works in few-shot class incremental learning (FSCIL) and active class selection (ACS) have achieved promising results but were tested in constrained setups. Therefore, in this paper, we combine ideas from FSCIL and ACS to develop a novel framework that can allow an autonomous agent to continually learn new objects by asking its users to label only a few of the most informative objects in the environment. To this end, we build on a state-of-the-art (SOTA) FSCIL model and extend it with techniques from ACS literature. We term this model Few-shot Incremental Active class SeleCtiOn (FIASco). We further integrate a potential field-based navigation technique with our model to develop a complete framework that can allow an agent to process and reason on its sensory data through the FIASco model, navigate towards the most informative object in the environment, gather data about the object through its sensors and incrementally update the FIASco model. Experimental results on a simulated agent and a real robot show the significance of our approach for long-term real-world robotics applications.} }
Endnote
%0 Conference Paper %T Active Class Selection for Few-Shot Class-Incremental Learning %A Christopher McClurg %A Ali Ayub %A Harsh Tyagi %A Sarah M. Rajtmajer %A Alan R. Wagner %B Proceedings of The 2nd Conference on Lifelong Learning Agents %C Proceedings of Machine Learning Research %D 2023 %E Sarath Chandar %E Razvan Pascanu %E Hanie Sedghi %E Doina Precup %F pmlr-v232-mcclurg23a %I PMLR %P 811--827 %U https://proceedings.mlr.press/v232/mcclurg23a.html %V 232 %X For real-world applications, robots will need to continually learn in their environments through limited interactions with their users. Toward this, previous works in few-shot class incremental learning (FSCIL) and active class selection (ACS) have achieved promising results but were tested in constrained setups. Therefore, in this paper, we combine ideas from FSCIL and ACS to develop a novel framework that can allow an autonomous agent to continually learn new objects by asking its users to label only a few of the most informative objects in the environment. To this end, we build on a state-of-the-art (SOTA) FSCIL model and extend it with techniques from ACS literature. We term this model Few-shot Incremental Active class SeleCtiOn (FIASco). We further integrate a potential field-based navigation technique with our model to develop a complete framework that can allow an agent to process and reason on its sensory data through the FIASco model, navigate towards the most informative object in the environment, gather data about the object through its sensors and incrementally update the FIASco model. Experimental results on a simulated agent and a real robot show the significance of our approach for long-term real-world robotics applications.
APA
McClurg, C., Ayub, A., Tyagi, H., Rajtmajer, S.M. & Wagner, A.R.. (2023). Active Class Selection for Few-Shot Class-Incremental Learning. Proceedings of The 2nd Conference on Lifelong Learning Agents, in Proceedings of Machine Learning Research 232:811-827 Available from https://proceedings.mlr.press/v232/mcclurg23a.html.

Related Material