Research map / Detection and segmentation / Open-set and few-shot detection
Benchmarking Unsupervised Object Representations for Video Sequences
Marissa A. Weis, Kashyap Chitta, Yash Sharma, Wieland Brendel, Matthias Bethge, Andreas Geiger, Alexander S. Ecker
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 |
|---|---|---|---|
| add temporal warm start frames before evaluation (ablation) | ViMON · VMDS | MOTA | Improved |
| add temporal warm start frames before evaluation (ablation) | OP3 · VMDS | MOTA | Improved |
| add temporal warm start frames before evaluation (ablation) | SCALOR · VMDS | MOTA | Hurt |
| add temporal warm start frames beyond 2-3 frames (ablation) | TBA · VMDS | MOTA | No clear effect |
| increase number of objects in the video | ViMON, TBA, SCALOR · VMDS | failure cases | Hurt |
| add heavy occlusion (challenge set) | all models · VMDS occlusion challenge set | MOTA | Hurt |
| make objects the same color | all models · VMDS same-color challenge set | MOTA | Hurt |
| use only large objects | TBA · VMDS large-object challenge set | MOTA | Hurt |
| add continuous rotation of objects (out-of-distribution) | ViMON · VMDS rotation o.o.d. | tracking performance | Hurt |
| add continuous size changes (out-of-distribution) | ViMON · VMDS size o.o.d. | tracking performance | Hurt |
| add continuous color changes (out-of-distribution) | all models · VMDS color o.o.d. | tracking performance | No clear effect |
| use textured ImageNet content for objects and background | ViMON, OP3, SCALOR, MONet · texVMDS | detection and tracking | Hurt |
Closest papers by meaning
- Exploiting Spatial Invariance for Scalable Unsupervised Object Tracking (AAAI 2020 · similarity 0.90)
- Learning Features by Watching Objects Move (CVPR 2017 · similarity 0.90)
- Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos (NeurIPS 2022 · similarity 0.90)
- Object Pursuit: Building a Space of Objects via Discriminative Weight Generation (ICLR 2022 · similarity 0.89)
- Object-Centric Multiple Object Tracking (ICCV 2023 · similarity 0.89)
- Temporally Consistent Object-Centric Learning by Contrasting Slots (CVPR 2025 · similarity 0.89)
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.
