Research map / 3D vision / 3D generation
Transformation-Grounded Image Generation Network for Novel 3D View Synthesis
Check 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 |
|---|---|---|---|
| Use transformation-grounded completion with a visibility map instead of AFN | Novel view synthesis · ShapeNet car | L1: .146 -> .133 | Helped |
| Use transformation-grounded completion with a visibility map instead of AFN | Novel view synthesis · ShapeNet chair | L1: .240 -> .230 | Helped |
| Add the visibility map to the completion pipeline (ablation) | TVSN · ShapeNet car | L1: .132 -> .133 | No clear effect |
| Add the visibility map to the completion pipeline (ablation) | TVSN · ShapeNet chair | SSIM: .895 -> .894 | No clear effect |
| Use a combined VGG16 perceptual and adversarial loss rather than L1 alone | Single-stage encoder-decoder · ShapeNet car | L1: .168 -> .194 | Hurt |
| Use a combined VGG16 perceptual and adversarial loss rather than L1 alone | Single-stage encoder-decoder · ShapeNet chair | L1: .248 -> .242 | Helped |
| Combine VGG16 perceptual and adversarial losses instead of either loss alone (ablation) | View completion network · Novel view synthesis | visual quality | Helped |
| Use random rather than pretrained loss-network weights (ablation) | Perceptual loss network · Novel view synthesis | perceptual quality | Hurt |
| Fine-tune both networks end-to-end after training completion with DOAFN fixed (ablation) | TVSN · Novel view synthesis | image quality | No clear effect |
Closest papers by meaning
- Visual Object Networks: Image Generation with Disentangled 3D Representations (NeurIPS 2018 · similarity 0.90)
- Towards Unsupervised Learning of Generative Models for 3D Controllable Image Synthesis (CVPR 2020 · similarity 0.89)
- Novel View Synthesis with Pixel-Space Diffusion Models (CVPR 2025 · similarity 0.88)
- CAD: Photorealistic 3D Generation via Adversarial Distillation (CVPR 2024 · similarity 0.87)
- EscherNet: A Generative Model for Scalable View Synthesis (CVPR 2024 · similarity 0.87)
- GenRec: Knowing Where to Reconstruct and Where to Generate (NeurIPS 2026 · similarity 0.87)
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.