Research map / Image generation / Autoregressive and latent generation
Deep Feature Interpolation for Image Content Changes
Paul Upchurch, Jacob Gardner, Geoff Pleiss, Robert Pless, Noah Snavely, Kavita Bala, Kilian Weinberger
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 DFI instead of AEGAN for attribute editing | LFW face attribute transformations · LFW, older | DFI-to-AEGAN preference ratio: 4.57:1 | Improved |
| Use DFI instead of AEGAN for attribute editing | LFW face attribute transformations · LFW, mouth open | DFI-to-AEGAN preference ratio: 7.09:1 | Improved |
| Use DFI instead of AEGAN for attribute editing | LFW face attribute transformations · LFW, eyes open | DFI-to-AEGAN preference ratio: 17.6:1 | Improved |
| Use DFI instead of AEGAN for attribute editing | LFW face attribute transformations · LFW, smiling | DFI-to-AEGAN preference ratio: 20.6:1 | Improved |
| Use DFI instead of AEGAN for attribute editing | LFW face attribute transformations · LFW, moustache | DFI-to-AEGAN preference ratio: 24.5:1 | Improved |
| Use DFI instead of AEGAN for attribute editing | LFW face attribute transformations · LFW, glasses | DFI-to-AEGAN preference ratio: 38.3:1 | Improved |
| Use DFI instead of disCVAE for inpainting when half the image is missing | DFI · LFW inpainting, half image missing | Inpainting quality | Hurt |
| Increase the number of source and target neighbors too far (ablation) | DFI inpainting · LFW inpainting | Visual plausibility | Hurt |
| Set the number of source and target neighbors too low (ablation) | DFI inpainting · LFW inpainting | Visual fit to surrounding pixels | Hurt |
| Use a transformation strength that is too small or too large (ablation) | DFI · Image transformations | Visual naturalness | Hurt |
| Increase transformation strength (ablation) | DFI · Image transformations | Transformation prominence | Improved |
Closest papers by meaning
- Video Frame Interpolation With Transformer (CVPR 2022 · similarity 0.86)
- Deep Network Interpolation for Continuous Imagery Effect Transition (CVPR 2019 · similarity 0.86)
- LDMVFI: Video Frame Interpolation with Latent Diffusion Models (AAAI 2024 · similarity 0.84)
- LookHere: Vision Transformers with Directed Attention Generalize and Extrapolate (NeurIPS 2024 · similarity 0.82)
- AID: Attention Interpolation of Text-to-Image Diffusion (NeurIPS 2024 · similarity 0.82)
- High-Resolution Frame Interpolation with Patch-based Cascaded Diffusion (AAAI 2025 · similarity 0.82)
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.
