Research map / Detection and segmentation / Open-set and few-shot detection
Reusable Slotwise Mechanisms
Trang Nguyen, Amin Mansouri, Kanika Madan, Nguyen Duy Khuong, Kartik Ahuja, Dianbo Liu, Yoshua Bengio
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 CCI to inform mechanism selection and slot updates | RSM · Video prediction on OBJ3D and CLEVRER | LPIPS×100 | Improved |
| Use CCI to inform mechanism selection and slot updates | RSM · CLEVRER | FG-mIoU: +3.05 points | Improved |
| Use RSM-p parallel mechanism assignment instead of sequential updates (ablation) | RSM · Video prediction | LPIPS | Mixed |
| Mask CCI during mechanism selection (ablation) | RSM · PHYRE object dynamics | mechanism assignment quality | Hurt |
| Mask CCI during slot updates (ablation) | RSM · PHYRE object dynamics | movement prediction quality | Hurt |
| Randomly assign mechanisms instead of selecting them with the learned distribution (ablation) | RSM · PHYRE object dynamics | movement prediction quality | Hurt |
| Increase the mechanism count substantially, including replicated mechanisms (ablation) | RSM · CLEVRER and Physion | task score | Hurt |
| Use a single higher-capacity mechanism rather than a set of specialized mechanisms (ablation) | RSM · CLEVRER and Physion | task score | Hurt |
| Use 1 or 2 hidden layers in the mechanism-selection network (ablation) | RSM · CLEVRER and Physion | task score | Improved |
| Allow blue objects to have moving-ball dynamics after training only on static blue objects (ablation) | RSM · PHYRE OOD dynamics | generalization | Improved |
Closest papers by meaning
- Slot State Space Models (NeurIPS 2024 · similarity 0.89)
- SOLD: Slot Object-Centric Latent Dynamics Models for Relational Manipulation Learning from Pixels (ICML 2025 · similarity 0.87)
- DoraemonGPT: Toward Understanding Dynamic Scenes with Large Language Models (Exemplified as A Video Agent) (ICML 2024 · similarity 0.85)
- PlaySlot: Learning Inverse Latent Dynamics for Controllable Object-Centric Video Prediction and Planning (ICML 2025 · similarity 0.85)
- Cycle Consistency Driven Object Discovery (ICLR 2024 · similarity 0.85)
- Parallelized Spatiotemporal Slot Binding for Videos (ICML 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.
