Research map / Detection and segmentation / Open-set and few-shot detection
Retinomorphic Object Detection in Asynchronous Visual Streams
Jianing Li, Xiao Wang, Lin Zhu, Jia Li, Tiejun Huang, Yonghong Tian
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 |
|---|---|---|---|
| Fuse Vidar spikes and DVS events using TAR and dynamic interaction | YOLOv3 retinomorphic detector · PKU-Vidar-DVS | mAP: 0.591 -> 0.647 | Improved |
| Fuse Vidar spikes and DVS events using TAR and dynamic interaction | YOLOv3 retinomorphic detector · KITTI simulated | mAP: 0.718 -> 0.762 | Improved |
| Use TAR-spikes instead of VTII (ablation) | Vidar YOLOv3 baseline · PKU-Vidar-DVS | mAP: 0.551 -> 0.579 | Improved |
| Add TAR-events to the TAR-spikes variant (ablation) | YOLOv3 retinomorphic detector · PKU-Vidar-DVS | mAP: 0.579 -> 0.606 | Improved |
| Add dynamic interaction fusion (ablation) | TAR-based YOLOv3 detector · PKU-Vidar-DVS | mAP: 0.606 -> 0.623 | Improved |
| Add ℓ1 regulation for channel sparsity (ablation) | TAR-based YOLOv3 detector with dynamic interaction · PKU-Vidar-DVS | mAP: 0.623 -> 0.647 | Improved |
| Replace dynamic interaction with typical fusion methods (ablation) | TAR-based YOLOv3 detector · PKU-Vidar-DVS | mAP: 0.647 versus 0.612; 0.647 versus 0.615; 0.647 versus 0.619; 0.647 versus 0.623; 0.647 versus 0.621; 0.647 versus 0.624; | Hurt |
| Increase temporal aggregation size from T=1 to T=3 (ablation) | Retinomorphic detector · PKU-Vidar-DVS | mAP: 0.623 -> 0.647 | Improved |
| Increase temporal aggregation size from T=3 to T=5 (ablation) | Retinomorphic detector · PKU-Vidar-DVS | mAP: 0.647 -> 0.649 | Improved |
| Increase temporal aggregation size from T=5 to T=7 or T=9 (ablation) | Retinomorphic detector · PKU-Vidar-DVS | mAP: 0.649; 0.650; 0.650 | No clear effect |
| Increase temporal aggregation size from T=3 to T=7 (ablation) | Retinomorphic detector · KITTI simulated | mAP: 0.762 -> 0.769 | Improved |
| Increase temporal aggregation size from T=7 to T=9 (ablation) | Retinomorphic detector · KITTI simulated | mAP: 0.769 -> 0.768 | Hurt |
Closest papers by meaning
- SODFormer: Streaming Object Detection With Transformer Using Events and Frames (TPAMI 2023 · similarity 0.89)
- AsyncBEV: Cross-modal flow alignment in Asynchronous 3D Object Detection (ICLR 2026 · similarity 0.88)
- Graph-Based Object Classification for Neuromorphic Vision Sensing (ICCV 2019 · similarity 0.88)
- Efficient Event-Based Object Detection: A Hybrid Neural Network with Spatial and Temporal Attention (CVPR 2025 · similarity 0.86)
- Finding Visual Saliency in Continuous Spike Stream (AAAI 2024 · similarity 0.86)
- SFOD: Spiking Fusion Object Detector (CVPR 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.
