Understanding Failures of Deep Networks via Robust Feature Extraction

• URL: http://arxiv.org/abs/2012.01750v2
• Date: Thu, 13 May 2021 12:44:02 GMT
• Title: Understanding Failures of Deep Networks via Robust Feature Extraction
• Authors: Sahil Singla, Besmira Nushi, Shital Shah, Ece Kamar, Eric Horvitz
• Abstract summary: We introduce and study a method aimed at characterizing and explaining failures by identifying visual attributes whose presence or absence results in poor performance. We leverage the representation of a separate robust model to extract interpretable features and then harness these features to identify failure modes.
• Score: 44.204907883776045
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Traditional evaluation metrics for learned models that report aggregate scores over a test set are insufficient for surfacing important and informative patterns of failure over features and instances. We introduce and study a method aimed at characterizing and explaining failures by identifying visual attributes whose presence or absence results in poor performance. In distinction to previous work that relies upon crowdsourced labels for visual attributes, we leverage the representation of a separate robust model to extract interpretable features and then harness these features to identify failure modes. We further propose a visualization method aimed at enabling humans to understand the meaning encoded in such features and we test the comprehensibility of the features. An evaluation of the methods on the ImageNet dataset demonstrates that: (i) the proposed workflow is effective for discovering important failure modes, (ii) the visualization techniques help humans to understand the extracted features, and (iii) the extracted insights can assist engineers with error analysis and debugging.

Related papers

• Unsupervised Model Diagnosis [49.36194740479798]
  This paper proposes Unsupervised Model Diagnosis (UMO) to produce semantic counterfactual explanations without any user guidance. Our approach identifies and visualizes changes in semantics, and then matches these changes to attributes from wide-ranging text sources.
  arXiv  Detail & Related papers  (2024-10-08T17:59:03Z)
• Spurious Feature Eraser: Stabilizing Test-Time Adaptation for Vision-Language Foundation Model [86.9619638550683]
  Vision-language foundation models have exhibited remarkable success across a multitude of downstream tasks due to their scalability on extensive image-text paired data. However, these models display significant limitations when applied to downstream tasks, such as fine-grained image classification, as a result of decision shortcuts''
  arXiv  Detail & Related papers  (2024-03-01T09:01:53Z)
• Vision-language Assisted Attribute Learning [53.60196963381315]
  Attribute labeling at large scale is typically incomplete and partial. Existing attribute learning methods often treat the missing labels as negative or simply ignore them all during training. We leverage the available vision-language knowledge to explicitly disclose the missing labels for enhancing model learning.
  arXiv  Detail & Related papers  (2023-12-12T06:45:19Z)
• Loss Functions in the Era of Semantic Segmentation: A Survey and Outlook [11.119967679567587]
  Loss functions are crucial for shaping the development of deep learning-based segmentation algorithms. We provide a novel taxonomy and review of how these loss functions are customized and leveraged in image segmentation. We conclude this review by identifying current challenges and unveiling future research opportunities.
  arXiv  Detail & Related papers  (2023-12-08T22:06:05Z)
• A Dual-Perspective Approach to Evaluating Feature Attribution Methods [43.16453263420591]
  We propose two new perspectives within the faithfulness paradigm that reveal intuitive properties: soundness and completeness. Soundness assesses the degree to which attributed features are truly predictive features, while completeness examines how well the resulting attribution reveals all the predictive features. We apply these metrics to mainstream attribution methods, offering a novel lens through which to analyze and compare feature attribution methods.
  arXiv  Detail & Related papers  (2023-08-17T12:41:04Z)
• Automated Deception Detection from Videos: Using End-to-End Learning Based High-Level Features and Classification Approaches [0.0]
  We propose a multimodal approach combining deep learning and discriminative models for deception detection. We employ convolutional end-to-end learning to analyze gaze, head pose, and facial expressions. Our approach is evaluated on five datasets, including a new Rolling-Dice Experiment motivated by economic factors.
  arXiv  Detail & Related papers  (2023-07-13T08:45:15Z)
• Data AUDIT: Identifying Attribute Utility- and Detectability-Induced Bias in Task Models [8.420252576694583]
  We present a first technique for the rigorous, quantitative screening of medical image datasets. Our method decomposes the risks associated with dataset attributes in terms of their detectability and utility. Using our method, we show our screening method reliably identifies nearly imperceptible bias-inducing artifacts.
  arXiv  Detail & Related papers  (2023-04-06T16:50:15Z)
• Information-Theoretic Odometry Learning [83.36195426897768]
  We propose a unified information theoretic framework for learning-motivated methods aimed at odometry estimation. The proposed framework provides an elegant tool for performance evaluation and understanding in information-theoretic language.
  arXiv  Detail & Related papers  (2022-03-11T02:37:35Z)
• A Simple and Effective Self-Supervised Contrastive Learning Framework for Aspect Detection [15.36713547251997]
  We propose a self-supervised contrastive learning framework and an attention-based model equipped with a novel smooth self-attention (SSA) module for the UAD task. Our methods outperform several recent unsupervised and weakly supervised approaches on publicly available benchmark user review datasets.
  arXiv  Detail & Related papers  (2020-09-18T22:13:49Z)
• Learning What Makes a Difference from Counterfactual Examples and Gradient Supervision [57.14468881854616]
  We propose an auxiliary training objective that improves the generalization capabilities of neural networks. We use pairs of minimally-different examples with different labels, a.k.a counterfactual or contrasting examples, which provide a signal indicative of the underlying causal structure of the task. Models trained with this technique demonstrate improved performance on out-of-distribution test sets.
  arXiv  Detail & Related papers  (2020-04-20T02:47:49Z)

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.