no code implementations • 2 Apr 2024 • Matthew Kowal, Richard P. Wildes, Konstantinos G. Derpanis
Understanding what deep network models capture in their learned representations is a fundamental challenge in computer vision.
no code implementations • 23 Jan 2024 • Shih-Han Chou, Matthew Kowal, Yasmin Niknam, Diana Moyano, Shayaan Mehdi, Richard Pito, Cheng Zhang, Ian Knopke, Sedef Akinli Kocak, Leonid Sigal, Yalda Mohsenzadeh
Towards a solution for designing this ability in algorithms, we present a large-scale analysis on an in-house dataset collected by the Reuters News Agency, called Reuters Video-Language News (ReutersViLNews) dataset which focuses on high-level video-language understanding with an emphasis on long-form news.
no code implementations • 19 Jan 2024 • Matthew Kowal, Achal Dave, Rares Ambrus, Adrien Gaidon, Konstantinos G. Derpanis, Pavel Tokmakov
Concretely, we seek to explain the decision-making process of video transformers based on high-level, spatiotemporal concepts that are automatically discovered.
no code implementations • 3 Nov 2022 • Matthew Kowal, Mennatullah Siam, Md Amirul Islam, Neil D. B. Bruce, Richard P. Wildes, Konstantinos G. Derpanis
(ii) Some datasets that are assumed to be biased toward dynamics are actually biased toward static information.
1 code implementation • CVPR 2022 • Matthew Kowal, Mennatullah Siam, Md Amirul Islam, Neil D. B. Bruce, Richard P. Wildes, Konstantinos G. Derpanis
To show the efficacy of our approach, we analyse two widely studied tasks, action recognition and video object segmentation.
no code implementations • 20 Oct 2021 • Md Amirul Islam, Matthew Kowal, Sen Jia, Konstantinos G. Derpanis, Neil D. B. Bruce
Extensive experiments demonstrate the high quality of our generated pseudo-labels and effectiveness of the proposed framework in a variety of domains.
no code implementations • 23 Aug 2021 • Md Amirul Islam, Matthew Kowal, Konstantinos G. Derpanis, Neil D. B. Bruce
The premise is based on the notion of feature binding, which is defined as the process by which activations spread across space and layers in the network are successfully integrated to arrive at a correct inference decision.
1 code implementation • ICCV 2021 • Md Amirul Islam, Matthew Kowal, Sen Jia, Konstantinos G. Derpanis, Neil D. B. Bruce
In this paper, we challenge the common assumption that collapsing the spatial dimensions of a 3D (spatial-channel) tensor in a convolutional neural network (CNN) into a vector via global pooling removes all spatial information.
no code implementations • 28 Jan 2021 • Md Amirul Islam, Matthew Kowal, Sen Jia, Konstantinos G. Derpanis, Neil D. B. Bruce
; (ii) Does position encoding affect the learning of semantic representations?
no code implementations • 27 Jan 2021 • Md Amirul Islam, Matthew Kowal, Patrick Esser, Sen Jia, Bjorn Ommer, Konstantinos G. Derpanis, Neil Bruce
Contrasting the previous evidence that neurons in the later layers of a Convolutional Neural Network (CNN) respond to complex object shapes, recent studies have shown that CNNs actually exhibit a `texture bias': given an image with both texture and shape cues (e. g., a stylized image), a CNN is biased towards predicting the category corresponding to the texture.
no code implementations • 1 Jan 2021 • Md Amirul Islam, Matthew Kowal, Sen Jia, Konstantinos G. Derpanis, Neil Bruce
Finally, we demonstrate the implications of these findings on a number of real-world tasks to show that position information can act as a feature or a bug.
no code implementations • ICLR 2021 • Md Amirul Islam, Matthew Kowal, Patrick Esser, Sen Jia, Björn Ommer, Konstantinos G. Derpanis, Neil Bruce
Contrasting the previous evidence that neurons in the later layers of a Convolutional Neural Network (CNN) respond to complex object shapes, recent studies have shown that CNNs actually exhibit a 'texture bias': given an image with both texture and shape cues (e. g., a stylized image), a CNN is biased towards predicting the category corresponding to the texture.
no code implementations • 13 Aug 2020 • Md Amirul Islam, Matthew Kowal, Konstantinos G. Derpanis, Neil D. B. Bruce
In this paper, we present a strategy for training convolutional neural networks to effectively resolve interference arising from competing hypotheses relating to inter-categorical information throughout the network.
no code implementations • 23 Aug 2019 • Matthew Kowal, Gillian Sandison, Len Yabuki-Soh, Raner la Bastide
Driver inattention is a large problem on the roads around the world.