Research map / Detection and segmentation / Open-set and few-shot detection
Context-aware Synthesis and Placement of Object Instances
Donghoon Lee, Sifei Liu, Jinwei Gu, Ming-Yu Liu, Ming-Hsuan Yang, Jan Kautz
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 all four discriminators rather than omit D_layout^box (ablation) | Context-aware synthesis and placement · Cityscapes person insertion | YOLOv3 recall: 0.60 -> 0.79 | Improved |
| use D_affine (ablation) | Context-aware synthesis and placement · Cityscapes person insertion | YOLOv3 recall: 0.70 -> 0.79 | Improved |
| use D_layout^instance (ablation) | Context-aware synthesis and placement · Cityscapes person insertion | YOLOv3 recall: 0.68 -> 0.79 | Improved |
| use D_shape (ablation) | Context-aware synthesis and placement · Cityscapes person insertion | YOLOv3 recall: 0.71 -> 0.79 | Improved |
| include input reconstruction loss (ablation) | Where and what modules · Object insertion; diverse locations and shapes | diversity | Improved |
| include supervised VAE-adversarial loss (ablation) | Where and what modules · Object insertion; diverse locations and shapes | diversity | Improved |
| predict where before generating what, instead of generating shape first | Joint location and shape generation · Object insertion | shape plausibility | Improved |
| decompose location and shape generation instead of generating an instance directly | Object insertion · Object insertion | shape plausibility | Improved |
Closest papers by meaning
- Add-it: Training-Free Object Insertion in Images With Pretrained Diffusion Models (ICLR 2025 · similarity 0.88)
- Learning Object Placement by Inpainting for Compositional Data Augmentation (ECCV 2020 · similarity 0.88)
- Example-Guided Image Synthesis using Masked Spatial-Channel Attention and Self-Supervision (ECCV 2020 · similarity 0.87)
- Click2Mask: Local Editing with Dynamic Mask Generation (AAAI 2025 · similarity 0.87)
- Improving Editability in Image Generation with Layer-wise Memory (CVPR 2025 · similarity 0.85)
- SmartMask: Context Aware High-Fidelity Mask Generation for Fine-grained Object Insertion and Layout Control (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.
