Research map / Graph learning / Knowledge and temporal graphs
An Actor-Centric Causality Graph for Asynchronous Temporal Inference in Group Activity
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 ACCG asynchronous causality relations to the synchronous base model | Base model with Inception-v3 · Volleyball | Group accuracy: 93.6 -> 95.0 | Helped |
| Add ACCG asynchronous causality relations to the synchronous base model | Base model with Inception-v3 · Volleyball | Individual accuracy: 83.8 -> 85.6 | Helped |
| Add ACCG asynchronous causality relations to the synchronous base model | Base model with Inception-v3 · Collective Activity | Group accuracy: 92.6 -> 94.5 | Helped |
| Add adaptive temporal shifting to synchronize cause and effect features (ablation) | ACCG · Volleyball | Group accuracy: 93.6 -> 95.0 | Helped |
| Add adaptive temporal shifting to synchronize cause and effect features (ablation) | ACCG · Volleyball | Individual accuracy: 85.2 -> 85.6 | Helped |
| Use four candidate delays {0,1,2,3} instead of {0,1,2} (ablation) | ACCG · Volleyball | Group accuracy: 95.0 -> 95.0 | No clear effect |
| Use four candidate delays {0,1,2,3} instead of {0,1,2} (ablation) | ACCG · Volleyball | Individual accuracy: 85.6 -> 85.6 | No clear effect |
| Include cause-action features in channel-wise fusion (ablation) | ACCG · Volleyball | Group accuracy: 93.3 -> 95.0 | Helped |
| Set the fusion channel ratio to d=6 rather than d=8 (ablation) | ACCG · Volleyball | Group accuracy: 93.5 -> 95.0 | Helped |
| Add graph attention to ACCG (ablation) | ACCG · Volleyball | Group accuracy: 93.3 -> 93.6 | Helped |
| Combine ARG with ACCG in two complementary branches (ablation) | ARG · Volleyball | Group accuracy: 92.5 -> 93.9 | Helped |
| Add ACCG without attention to the base model (ablation) | Base model · Volleyball | Group accuracy: 93.6 -> 94.3 | Helped |
Closest papers by meaning
- Sim-to-Real Causal Transfer: A Metric Learning Approach to Causally-Aware Interaction Representations (CVPR 2025 · similarity 0.85)
- Bi-Causal: Group Activity Recognition via Bidirectional Causality (CVPR 2024 · similarity 0.84)
- Rethinking Causal Relationships Learning in Graph Neural Networks (AAAI 2024 · similarity 0.84)
- Causal Representation Learning Made Identifiable by Grouping of Observational Variables (ICML 2024 · similarity 0.84)
- Local Identifying Causal Relations in the Presence of Latent Variables (ICML 2025 · similarity 0.84)
- UnCLe: Towards Scalable Dynamic Causal Discovery in Non-linear Temporal Systems (NeurIPS 2025 · similarity 0.84)
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.