Research map / Video / Video generation
Pix2Gif: Motion-Guided Diffusion for GIF Generation
Hitesh Kandala, Jianfeng Gao, Jianwei Yang
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 |
|---|---|---|---|
| train InstructPix2Pix on the curated TGIF data with a flow phrase appended (Pix2Gif-Base) (ablation) | InstructPix2Pix · Motion coherency, cfg_img=1.4 | L2: 23.429 -> 7.580 | Improved |
| add motion embedding layer, warping module and perceptual loss to Pix2Gif-Base (ablation) | Pix2Gif-Base · Motion coherency, cfg_img=1.4 | L2: 7.580 -> 1.746 | Improved |
| add motion embedding layer to encode motion magnitude (ablation) | Pix2Gif-Base · image-video similarity (100 samples) | source/target frame similarity score | Improved |
| add warping module whose output replaces source feature as LDM input (Pix2Gif-Warp) (ablation) | Pix2Gif · image-video similarity (100 samples) | average similarity score | Hurt |
| concatenate warped feature with source image feature instead of adding (ablation) | Pix2Gif · image-video similarity, cfg_img in [1.6, 2.2] | average similarity score | Improved |
| add perceptual loss on warped latent (ablation) | Pix2Gif-Warp-concat · image-video similarity, cfg_img in [1.6, 2.2] | motion consistency vs extent of motion | Mixed |
| Pix2Gif versus I2VGen-XL and DynamiCrafter on CLIPSim | image-text-to-video generation · UCF-101 and MSR-VTT | CLIPSim: UCF-101: 0.2865 -> 0.2815 | Hurt |
| frame-by-frame image translation instead of full-sequence 3D video diffusion | image-to-GIF generation · UCF-101 zero-shot | FVD: 527.06 -> 285.02 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DynamiCrafter — DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors (ECCV 2024) · better on 3 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- MeDM: Mediating Image Diffusion Models for Video-to-Video Translation with Temporal Correspondence Guidance (AAAI 2024 · similarity 0.88)
- Optical-Flow Guided Prompt Optimization for Coherent Video Generation (CVPR 2025 · similarity 0.87)
- LLM-grounded Video Diffusion Models (ICLR 2024 · similarity 0.87)
- Grid Diffusion Models for Text-to-Video Generation (CVPR 2024 · similarity 0.87)
- Make Pixels Dance: High-Dynamic Video Generation (CVPR 2024 · similarity 0.87)
- Hierarchical Spatio-temporal Decoupling for Text-to-Video Generation (CVPR 2024 · 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.
