MDrive benchmark shows multi-agent cooperative driving systems generally outperform single-agent ones in closed-loop settings but perception sharing does not always improve planning and negotiation can harm performance in complex traffic.
”CoBEVT: Cooperative bird’s eye view semantic segmentation with sparse transformers.” arXiv preprint arXiv:2207.02202 (2022)
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 5roles
background 1polarities
background 1representative citing papers
A new online attack framework manipulates object poses in shared CAV perception data below detection thresholds, propagating errors to cause unsafe trajectory predictions and behaviors in up to 50% of tested scenarios while evading defenses.
SHLE is a stereo pipeline with device tracking and temporal depth filtering that estimates height limits with under 10cm average error at 70m distance on the new Disparity Height dataset.
VQSOP applies sparsity-exploiting vector quantization and a dual-branch refinement module to cut communication volume by up to 82x while claiming state-of-the-art 3D occupancy prediction performance.
CooperDrive augments autonomous vehicle perception by sharing object-level data from BEV features, enabling earlier conflict anticipation and safer planning with 90 kbps bandwidth and 89 ms latency in real-world NLOS tests.
citing papers explorer
-
MDrive: Benchmarking Closed-Loop Cooperative Driving for End-to-End Multi-agent Systems
MDrive benchmark shows multi-agent cooperative driving systems generally outperform single-agent ones in closed-loop settings but perception sharing does not always improve planning and negotiation can harm performance in complex traffic.
-
From Stealthy Data Fabrication to Unsafe Driving: Realistic Scenario Attacks on Collaborative Perception
A new online attack framework manipulates object poses in shared CAV perception data below detection thresholds, propagating errors to cause unsafe trajectory predictions and behaviors in up to 50% of tested scenarios while evading defenses.
-
SHLE: Devices Tracking and Depth Filtering for Stereo-based Height Limit Estimation
SHLE is a stereo pipeline with device tracking and temporal depth filtering that estimates height limits with under 10cm average error at 70m distance on the new Disparity Height dataset.
-
Sparse-Aware Vector Quantization for Bandwidth-Efficient Collaborative 3D Semantic Occupancy Prediction
VQSOP applies sparsity-exploiting vector quantization and a dual-branch refinement module to cut communication volume by up to 82x while claiming state-of-the-art 3D occupancy prediction performance.
-
CooperDrive: Enhancing Driving Decisions Through Cooperative Perception
CooperDrive augments autonomous vehicle perception by sharing object-level data from BEV features, enabling earlier conflict anticipation and safer planning with 90 kbps bandwidth and 89 ms latency in real-world NLOS tests.