Understanding Video Transformers via Universal Concept Discovery

• URL: http://arxiv.org/abs/2401.10831v3
• Date: Wed, 10 Apr 2024 15:19:07 GMT
• Title: Understanding Video Transformers via Universal Concept Discovery
• Authors: Matthew Kowal, Achal Dave, Rares Ambrus, Adrien Gaidon, Konstantinos G. Derpanis, Pavel Tokmakov,
• Abstract summary: We seek to explain the decision-making process of transformers based on high-level,temporal concepts that are automatically discovered. We introduce the first Video Transformer Concept Discovery (VTCD) algorithm. The resulting concepts are highly interpretable, revealingtemporal reasoning mechanisms and object-centric representations in unstructured video models.
• Score: 44.869479587300525
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: This paper studies the problem of concept-based interpretability of transformer representations for videos. Concretely, we seek to explain the decision-making process of video transformers based on high-level, spatiotemporal concepts that are automatically discovered. Prior research on concept-based interpretability has concentrated solely on image-level tasks. Comparatively, video models deal with the added temporal dimension, increasing complexity and posing challenges in identifying dynamic concepts over time. In this work, we systematically address these challenges by introducing the first Video Transformer Concept Discovery (VTCD) algorithm. To this end, we propose an efficient approach for unsupervised identification of units of video transformer representations - concepts, and ranking their importance to the output of a model. The resulting concepts are highly interpretable, revealing spatio-temporal reasoning mechanisms and object-centric representations in unstructured video models. Performing this analysis jointly over a diverse set of supervised and self-supervised representations, we discover that some of these mechanism are universal in video transformers. Finally, we show that VTCD can be used for fine-grained action recognition and video object segmentation.

Related papers

• Object-Centric Temporal Consistency via Conditional Autoregressive Inductive Biases [69.46487306858789]
  Conditional Autoregressive Slot Attention (CA-SA) is a framework that enhances the temporal consistency of extracted object-centric representations in video-centric vision tasks. We present qualitative and quantitative results showing that our proposed method outperforms the considered baselines on downstream tasks.
  arXiv  Detail & Related papers  (2024-10-21T07:44:44Z)
• Introduction to Transformers: an NLP Perspective [59.0241868728732]
  We introduce basic concepts of Transformers and present key techniques that form the recent advances of these models. This includes a description of the standard Transformer architecture, a series of model refinements, and common applications.
  arXiv  Detail & Related papers  (2023-11-29T13:51:04Z)
• Understanding Video Transformers for Segmentation: A Survey of Application and Interpretability [10.180033230324561]
  Recently, approaches in this research area shifted from concentrating on ConvNet-based to transformer-based models. Various interpretability approaches have appeared for transformer models and video temporal dynamics.
  arXiv  Detail & Related papers  (2023-10-18T19:58:25Z)
• Holistically Explainable Vision Transformers [136.27303006772294]
  We propose B-cos transformers, which inherently provide holistic explanations for their decisions. Specifically, we formulate each model component - such as the multi-layer perceptrons, attention layers, and the tokenisation module - to be dynamic linear. We apply our proposed design to Vision Transformers (ViTs) and show that the resulting models, dubbed Bcos-ViTs, are highly interpretable and perform competitively to baseline ViTs.
  arXiv  Detail & Related papers  (2023-01-20T16:45:34Z)
• Hierarchical Multimodal Transformer to Summarize Videos [103.47766795086206]
  Motivated by the great success of transformer and the natural structure of video (frame-shot-video), a hierarchical transformer is developed for video summarization. To integrate the two kinds of information, they are encoded in a two-stream scheme, and a multimodal fusion mechanism is developed based on the hierarchical transformer. Practically, extensive experiments show that HMT surpasses most of the traditional, RNN-based and attention-based video summarization methods.
  arXiv  Detail & Related papers  (2021-09-22T07:38:59Z)
• Transformers in Vision: A Survey [101.07348618962111]
  Transformers enable modeling long dependencies between input sequence elements and support parallel processing of sequence. Transformers require minimal inductive biases for their design and are naturally suited as set-functions. This survey aims to provide a comprehensive overview of the Transformer models in the computer vision discipline.
  arXiv  Detail & Related papers  (2021-01-04T18:57:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.