Research map / Architectures and efficiency / RNNs and normalization
CERN: Confidence-Energy Recurrent Network for Group Activity Recognition
Check 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 confidence-regularized energy layer to CERN-1 (ablation) | CERN-1 · Collective Activity | MCA: 83.8 -> 84.8 | Helped |
| Add the confidence-regularized energy layer to CERN-1 (ablation) | CERN-1 · Volleyball (1 group) | MCA: 33.3 -> 34.4 | Mixed |
| Add p-value regularization to CERN-2 (ablation) | CERN-2 · Collective Activity | MCA: 83.8 -> 87.2 | Helped |
| Add p-value regularization to CERN-2 (ablation) | CERN-2 · Volleyball (1 group) | MCA: 71.7 -> 73.5 | Helped |
| Add p-value regularization to CERN-2 (ablation) | CERN-2 · Volleyball (2 groups) | MCA: 82.2 -> 83.3 | Helped |
| Add an event LSTM for group-activity prediction (ablation) | CERN-1, yielding CERN-2 · Vollyball (1 group) | MCA: 34.4 -> 73.5 | Helped |
| Use confidence p-values in inference (ablation) | CERN-2 · Collective Activity with corrupted trajectories | group activity accuracy degradation | Helped |
| Use confidence p-values in inference (ablation) | CERN-2 · Volleyball with distorted inputs | group activity accuracy degradation | Helped |
Closest papers by meaning
- A Hierarchical Deep Temporal Model for Group Activity Recognition (CVPR 2016 · similarity 0.89)
- Learning Activity Progression in LSTMs for Activity Detection and Early Detection (CVPR 2016 · similarity 0.87)
- Structure Inference Machines: Recurrent Neural Networks for Analyzing Relations in Group Activity Recognition (CVPR 2016 · similarity 0.87)
- Recurrent Modeling of Interaction Context for Collective Activity Recognition (CVPR 2017 · similarity 0.86)
- Rep Deep & Machine Learning: Exemplar-Free Continual Video Action Recognition via Slow-Fast Collaborative Learning (AAAI 2026 · similarity 0.85)
- HARDVS: Revisiting Human Activity Recognition with Dynamic Vision Sensors (AAAI 2024 · 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.