Research map / Adversarial robustness and security / Attacks on text and graphs
Learning Mutual View Information Graph for Adaptive Adversarial Collaborative Perception
Yihang Tao, Senkang Hu, Haonan An, Zhengru Fang, Hangcheng Cao, Yuguang Fang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| use MVIG spoof attack instead of RC (spoof) against CAD | MVIG attack · Adv-OPV2V spoofing, CAD defense | DSR: 83.5 -> 32.0 | Improved |
| use MVIG removal attack instead of RC (remove) against CAD | MVIG attack · Adv-OPV2V removal, CAD defense | DSR: 90.1 -> 78.2 | Improved |
| use MVIG spoof instead of RC (spoof) against ROBOSAC | MVIG attack · Adv-OPV2V spoofing, ROBOSAC defense | DSR: 12.0 -> 14.8 | Hurt |
| use MVIG spoof instead of RC (spoof) against GCP | MVIG attack · Adv-OPV2V spoofing, GCP defense | DSR: 12.2 -> 13.0 | Hurt |
| use MVIG spoof instead of RC (spoof) against CP-Guard | MVIG attack · Adv-OPV2V spoofing, CP-Guard defense | DSR: 18.1 -> 17.2 | No clear effect |
| use MVIG removal instead of RC (remove) against CP-Guard | MVIG attack · Adv-OPV2V removal, CP-Guard defense | DSR: 21.4 -> 21.3 | No clear effect |
| extend attack to 3-frame persistence with MVIG vs RC against CAD | MVIG attack · Adv-OPV2V spoofing, CAD defense, 3-frame persistence | DSR: 100.0 -> 63.2 | Improved |
| increase persistence from 0-frame to 3-frame for RC (spoof) | RC (spoof) baseline · CAD defense | DSR: 83.5 -> 100.0 | Hurt |
| enlarge impacted zone from 15m to 50m for spoof attack (ablation) | MVIG (spoof) · CAD defense | DSR (%): 40.2% -> 14.8% | Improved |
| enlarge impacted zone for removal attack (ablation) | MVIG (remove) · CAD defense | DSR (%): between 72.1% and 84.5% | No clear effect |
| add attack continuation threshold (0.35 to 0.55) instead of none (ablation) | MVIG (spoof) · CAD defense | DSR (%): 85.3% -> lower | Improved |
| use MVIG instead of RC at higher persistence against ROBOSAC+ for removal | MVIG (remove) · ROBOSAC+ 1-frame persistence | DSR: 28.1 -> 32.5 | Hurt |
Closest papers by meaning
- CP-Guard: Malicious Agent Detection and Defense in Collaborative Bird’s Eye View Perception (AAAI 2025 · similarity 0.87)
- RCP-Bench: Benchmarking Robustness for Collaborative Perception Under Diverse Corruptions (CVPR 2025 · similarity 0.85)
- Privacy-Preserving V2X Collaborative Perception Integrating Unknown Collaborators (AAAI 2025 · similarity 0.85)
- DSRC: Learning Density-Insensitive and Semantic-Aware Collaborative Representation Against Corruptions (AAAI 2025 · similarity 0.84)
- ReaLM: A Unified Red-Teaming Benchmark for Physical-World VLMs (NeurIPS 2026 · similarity 0.84)
- Value at Adversarial Risk: A Graph Defense Strategy against Cost-Aware Attacks (AAAI 2024 · 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.
