Research map / Video / Video generation
Flowception: Temporally Expansive Flow Matching for Video Generation
Tariq Berrada Ifriqi, John Nguyen, Karteek Alahari, Jakob Verbeek, Ricky T. Q. Chen
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 |
|---|---|---|---|
| replace data-independent frame insertion (random) with learned insertion rates (ablation) | Flowception · RealEstate10K I2V | FVD: 25.03 -> 21.80 | Improved |
| replace left-to-right insertion with learned insertion rates (ablation) | Flowception · RealEstate10K I2V | FVD: 23.61 -> 21.80 | Improved |
| replace causal with non-causal attention in the AR baseline (ablation) | AR baseline · RealEstate10K I2V | FVD: 47.48 -> 45.13 | Improved |
| use Flowception instead of autoregressive generation | Flowception · Kinetics-600 I2V | FVD: 201.34 -> 164.73 | Improved |
| use Flowception instead of full-sequence generation | Flowception · Tai-Chi-HD I2V | FVD: 27.30 -> 25.21 | No clear effect |
| use Flowception instead of full-sequence generation on LTX-2b T2V | LTX-2b · T2V proprietary | VBench imaging quality: 47.96 -> 51.37 | Improved |
| use Flowception instead of full-sequence generation on LTX-2b T2V | LTX-2b · T2V proprietary | VBench dynamic degree: 53.71 -> 52.43 | Hurt |
| use Flowception sampling instead of full-sequence sampling | LTX-2b · T2V sampling speed | wall-clock time: ~30% faster | Improved |
| use local attention windows instead of full attention (ablation) | Flowception vs full-sequence · RealEstate10K I2V | FVD: see Figure 8 (no numbers) | Mixed |
Closest papers by meaning
- FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis (CVPR 2024 · similarity 0.90)
- Contextual Flow Matching: Adaptive Step Selection in Flow Models for Efficient Visual Generation (NeurIPS 2026 · similarity 0.90)
- Pyramidal Flow Matching for Efficient Video Generative Modeling (ICLR 2025 · similarity 0.88)
- Optical-Flow Guided Prompt Optimization for Coherent Video Generation (CVPR 2025 · similarity 0.88)
- StreamingT2V: Consistent, Dynamic, and Extendable Long Video Generation from Text (CVPR 2025 · similarity 0.87)
- FlowMo: Variance-Based Flow Guidance for Coherent Motion in Video Generation (NeurIPS 2025 · 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.
