Exploring Foveation and Saccade for Improved Weakly-Supervised Localization

Timur Ibrayev, Manish Nagaraj, Amitangshu Mukherjee, Kaushik Roy
Proceedings of The 2nd Gaze Meets ML workshop, PMLR 226:61-89, 2024.

Abstract

Deep neural networks have become the de facto choice as feature extraction engines, ubiquitously used for computer vision tasks. The current approach is to process every input with uniform resolution in a one-shot manner and make all of the predictions at once. However, human vision is an "active" process that not only actively switches from one focus point to another within the visual field, but also applies spatially varying attention centered at such focus points. To bridge the gap, we propose incorporating the bio-plausible mechanisms of foveation and saccades to build an active object localization framework. While foveation enables it to process different regions of the input with variable degrees of detail, saccades allow it to change the focus point of such foveated regions. Our experiments show that these mechanisms improve the quality of predicted bounding boxes by capturing all the essential object parts while minimizing unnecessary background clutter. Additionally, they enable the resiliency of the method by allowing it to detect multiple objects while being trained only on data containing a single object per image. Finally, we explore the alignment of our method with human perception using the interesting "duck-rabbit" optical illusion. The code is available at: https://github.com/TimurIbrayev/FALcon.

Cite this Paper

BibTeX
@InProceedings{pmlr-v226-ibrayev24a, title = {Exploring Foveation and Saccade for Improved Weakly-Supervised Localization}, author = {Ibrayev, Timur and Nagaraj, Manish and Mukherjee, Amitangshu and Roy, Kaushik}, booktitle = {Proceedings of The 2nd Gaze Meets ML workshop}, pages = {61--89}, year = {2024}, editor = {Madu Blessing, Amarachi and Wu, Joy and Zanca, Dario and Krupinski, Elizabeth and Kashyap, Satyananda and Karargyris, Alexandros}, volume = {226}, series = {Proceedings of Machine Learning Research}, month = {16 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v226/ibrayev24a/ibrayev24a.pdf}, url = {https://proceedings.mlr.press/v226/ibrayev24a.html}, abstract = {Deep neural networks have become the de facto choice as feature extraction engines, ubiquitously used for computer vision tasks. The current approach is to process every input with uniform resolution in a one-shot manner and make all of the predictions at once. However, human vision is an "active" process that not only actively switches from one focus point to another within the visual field, but also applies spatially varying attention centered at such focus points. To bridge the gap, we propose incorporating the bio-plausible mechanisms of foveation and saccades to build an active object localization framework. While foveation enables it to process different regions of the input with variable degrees of detail, saccades allow it to change the focus point of such foveated regions. Our experiments show that these mechanisms improve the quality of predicted bounding boxes by capturing all the essential object parts while minimizing unnecessary background clutter. Additionally, they enable the resiliency of the method by allowing it to detect multiple objects while being trained only on data containing a single object per image. Finally, we explore the alignment of our method with human perception using the interesting "duck-rabbit" optical illusion. The code is available at: https://github.com/TimurIbrayev/FALcon.} }
Endnote
%0 Conference Paper %T Exploring Foveation and Saccade for Improved Weakly-Supervised Localization %A Timur Ibrayev %A Manish Nagaraj %A Amitangshu Mukherjee %A Kaushik Roy %B Proceedings of The 2nd Gaze Meets ML workshop %C Proceedings of Machine Learning Research %D 2024 %E Amarachi Madu Blessing %E Joy Wu %E Dario Zanca %E Elizabeth Krupinski %E Satyananda Kashyap %E Alexandros Karargyris %F pmlr-v226-ibrayev24a %I PMLR %P 61--89 %U https://proceedings.mlr.press/v226/ibrayev24a.html %V 226 %X Deep neural networks have become the de facto choice as feature extraction engines, ubiquitously used for computer vision tasks. The current approach is to process every input with uniform resolution in a one-shot manner and make all of the predictions at once. However, human vision is an "active" process that not only actively switches from one focus point to another within the visual field, but also applies spatially varying attention centered at such focus points. To bridge the gap, we propose incorporating the bio-plausible mechanisms of foveation and saccades to build an active object localization framework. While foveation enables it to process different regions of the input with variable degrees of detail, saccades allow it to change the focus point of such foveated regions. Our experiments show that these mechanisms improve the quality of predicted bounding boxes by capturing all the essential object parts while minimizing unnecessary background clutter. Additionally, they enable the resiliency of the method by allowing it to detect multiple objects while being trained only on data containing a single object per image. Finally, we explore the alignment of our method with human perception using the interesting "duck-rabbit" optical illusion. The code is available at: https://github.com/TimurIbrayev/FALcon.
APA
Ibrayev, T., Nagaraj, M., Mukherjee, A. & Roy, K.. (2024). Exploring Foveation and Saccade for Improved Weakly-Supervised Localization. Proceedings of The 2nd Gaze Meets ML workshop, in Proceedings of Machine Learning Research 226:61-89 Available from https://proceedings.mlr.press/v226/ibrayev24a.html.

Related Material