Research map / Detection and segmentation / CNN features
Video Object Detection with Locally-Weighted Deformable Neighbors
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 naive memory propagation between keyframes (ablation) | LWDN · ImageNet VID val | mAP (%): 73.0 -> 75.5 | Helped |
| add quality-aware memory aggregation (ablation) | LWDN · ImageNet VID val | mAP (%): 75.5 -> 76.3 | Helped |
| increase LWDN kernel size from 3 to 9 (ablation) | LWDN · ImageNet VID val | mAP (%): 74.7 -> 76.3 | Helped |
| increase LWDN kernel size from 9 to 11 (ablation) | LWDN · ImageNet VID val | mAP (%): 76.3 -> 75.7 | Hurt |
| set training keyframe interval to 10 instead of 4 (ablation) | LWDN · ImageNet VID val | mAP (%): 74.7 -> 76.3 | Helped |
| lengthen training keyframe interval from 10 to 12 (ablation) | LWDN · ImageNet VID val | mAP (%): 76.3 -> 75.4 | Hurt |
| lengthen testing keyframe interval from 6 to 14 (ablation) | LWDN · ImageNet VID val | mAP (%): 76.7 -> 75.2 | Hurt |
| raise memory gate g from 0.0 to 1.0 (ablation) | LWDN · ImageNet VID val | mAP (%): 75.2 -> 76.3 | Helped |
| replace per-frame R-FCN with LWDN video detection | R-FCN ResNet-101 · ImageNet VID val | mAP (%): 73.6 -> 76.3 | Helped |
Closest papers by meaning
- LSTFE-Net:Long Short-Term Feature Enhancement Network for Video Small Object Detection (CVPR 2023 · similarity 0.86)
- Mobile Video Object Detection With Temporally-Aware Feature Maps (CVPR 2018 · similarity 0.85)
- Memory Enhanced Global-Local Aggregation for Video Object Detection (CVPR 2020 · similarity 0.85)
- Object Detection From Video Tubelets With Convolutional Neural Networks (CVPR 2016 · similarity 0.85)
- TGBFormer: Transformer-GraphFormer Blender Network for Video Object Detection (AAAI 2025 · similarity 0.84)
- LiVOS: Light Video Object Segmentation with Gated Linear Matching (CVPR 2025 · similarity 0.83)
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.