Research map / Video / Action recognition
Spatial-Temporal Concept Based Explanation of 3D ConvNets
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add the 5 highest-score concepts rather than 5 randomly selected concepts | r3d · 10-class Kinetics-700, adding concepts | accuracy (%): 41.25 -> 46.67 | Helped |
| Add the 5 highest-score concepts rather than the 5 lowest-score concepts | r3d · 10-class Kinetics-700, adding concepts | accuracy (%): 41.46 -> 46.67 | Helped |
| Add the 5 highest-score concepts rather than 5 randomly selected concepts | i3d · 10-class Kinetics-700, adding concepts | accuracy (%): 56.67 -> 61.88 | Helped |
| Add the 5 highest-score concepts rather than the 5 lowest-score concepts | i3d · 10-class Kinetics-700, adding concepts | accuracy (%): 55.83 -> 61.88 | Helped |
| Remove the 5 highest-score concepts rather than 5 randomly selected concepts | r3d · 10-class Kinetics-700, removing concepts | accuracy (%): 28.13 -> 24.38 | Hurt |
| Remove the 5 highest-score concepts rather than the 5 lowest-score concepts | r3d · 10-class Kinetics-700, removing concepts | accuracy (%): 28.13 -> 24.38 | Hurt |
| Remove the 5 highest-score concepts rather than 5 randomly selected concepts | i3d · 10-class Kinetics-700, removing concepts | accuracy (%): 43.96 -> 41.88 | Hurt |
| Remove the 5 highest-score concepts rather than the 5 lowest-score concepts | i3d · 10-class Kinetics-700, removing concepts | accuracy (%): 46.25 -> 41.88 | Hurt |
Closest papers by meaning
- StNet: Local and Global Spatial-Temporal Modeling for Action Recognition (AAAI 2019 · similarity 0.87)
- Show and Tell: Visually Explainable Deep Neural Nets via Spatially-Aware Concept Bottleneck Models (CVPR 2025 · similarity 0.86)
- Comparing the Decision-Making Mechanisms by Transformers and CNNs via Explanation Methods (CVPR 2024 · similarity 0.85)
- PICNN: A Pathway towards Interpretable Convolutional Neural Networks (AAAI 2024 · similarity 0.85)
- Representational Similarity via Interpretable Visual Concepts (ICLR 2025 · similarity 0.85)
- GPT4Scene: Understand 3D Scenes from Videos with Vision-Language Models (ICLR 2026 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.