Research map / Detection and segmentation / Open-set and few-shot detection
Learning to Manipulate Individual Objects in an Image
Yanchao Yang, Yutong Chen, Stefano Soatto
Check 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 object-centric spatial disentanglement instead of MONet's segmentation and representation loss | object segmentation · Multi-dSprites | mean IoU: 0.84\pm6.4\delta -> 0.92\pm6.6\delta | Improved |
| Use object-centric spatial disentanglement instead of MONet's segmentation and representation loss | object segmentation · Objects Room | mean IoU: 0.80\pm8.3\delta -> 0.85\pm5.6\delta | Improved |
| Use object-centric spatial disentanglement instead of MONet's segmentation and representation loss | object segmentation · Multi-Texture | mean IoU: 0.37\pm0.3\delta -> 0.88\pm2.6\delta | Improved |
| Use object-centric spatial disentanglement instead of MONet's segmentation and representation loss | object segmentation · Flying Animals | mean IoU: 0.18\pm2.8\delta -> 0.81\pm5.5\delta | Improved |
| Add perceptual cycle-consistency (ablation) | object-centric segmentation · Multi-Texture sequences | identity switches: 21/255 -> 0/255 | Improved |
| Add contextual information separation based on edge-conditioned inpainting | object-centric segmentation · Multi-Texture | segmentation quality | Improved |
| Increase appearance and shape complexity while retaining the beta-VAE bottleneck | object-centric representation · Flying Animals | reconstruction detail and statistical disentanglement | Mixed |
Closest papers by meaning
- Unsupervised Robust Disentangling of Latent Characteristics for Image Synthesis (ICCV 2019 · similarity 0.88)
- GENESIS: Generative Scene Inference and Sampling with Object-Centric Latent Representations (ICLR 2020 · similarity 0.88)
- Thinking Outside the BBox: Unconstrained Generative Object Compositing (ECCV 2024 · similarity 0.87)
- An Image is Worth More Than a Thousand Words: Towards Disentanglement in The Wild (NeurIPS 2021 · similarity 0.87)
- Object centric architectures enable efficient causal representation learning (ICLR 2024 · similarity 0.87)
- Entity-Centric Reinforcement Learning for Object Manipulation from Pixels (ICLR 2024 · similarity 0.86)
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.
