Video generation: research map
2,177 papers on Video generation in Video, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 10 approaches. The busiest year so far is 2026.
What it is
Generating video with diffusion models: from text, from an image of a person (animation), by editing an existing video, or long videos generated piece by piece. Keeping motion realistic and objects consistent over time is the central challenge.
Terms you will see on this page:
- Text-to-video
- generating a video clip from a text prompt.
- Temporal consistency
- objects and appearance staying coherent across frames.
- World model (video)
- a video generator used to simulate how a scene evolves under actions.
Start reading with the most cited papers here:
- MoCoGAN: Decomposing Motion and Content for Video Generation (CVPR 2018 · 1,116 citations)
- One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing (CVPR 2021 · 445 citations)
Within Video, its share grew from 22.1% in 2023–24 to 44.5% in 2025–26 (228 → 1,121 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Video generation in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Trajectory-guided motion control rising: 6% → 15% of the topic (13 → 167 papers)
- World-model video generation rising: 1% → 9% of the topic (3 → 105 papers)
- Autoregressive long video generation rising: 6% → 12% of the topic (13 → 130 papers)
- Text-to-motion diffusion falling: 19% → 8% of the topic (43 → 95 papers)
- Diffusion-based video editing falling: 14% → 5% of the topic (32 → 58 papers)
Papers from 2025–2026 that are already cited most:
- DriveDreamer-2: LLM-Enhanced World Models for Diverse Driving Video Generation (AAAI 2025 · 59 citations)
- DepthCrafter: Generating Consistent Long Depth Sequences for Open-world Videos (CVPR 2025 · 51 citations)
- DriveDreamer4D: World Models Are Effective Data Machines for 4D Driving Scene Representation (CVPR 2025 · 35 citations)
- Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer (CVPR 2025 · 29 citations)
Approaches and key papers
Text-to-video generation · 1,356 papers
Keywords: video generation, video diffusion, text video
Approaches in this cluster:
- Efficient video diffusion models (398 papers)
Builds video diffusion models with latent decomposition, autoregressive distillation and training-free long-video generation. - Compositional text-to-video diffusion (231 papers)
Generates videos from text with improved motion, composition and extendable length, plus benchmarks. - Trajectory-guided motion control (322 papers)
Controls video generation motion with sparse trajectories, latent trajectory guidance and tracking signals. - Autoregressive long video generation (224 papers)
Generates long videos autoregressively with continuous tokens, propagation and rich perception, plus evaluation suites. - World-model video generation (181 papers)
Uses video generators as interactive world simulators and benchmarks their physical consistency.
Most cited and most cited since 2024:
- MoCoGAN: Decomposing Motion and Content for Video Generation (CVPR 2018 · 1,116 citations)
- VBench: Comprehensive Benchmark Suite for Video Generative Models (CVPR 2024 · 351 citations)
- VideoCrafter2: Overcoming Data Limitations for High-Quality Video Diffusion Models (CVPR 2024 · 223 citations)
- Follow Your Pose: Pose-Guided Text-to-Video Generation Using Pose-Free Videos (AAAI 2024 · 122 citations)
Animation, editing and audio · 821 papers
Keywords: audio, editing, talking
Approaches in this cluster:
- Diffusion-based human image animation (168 papers)
Animates portraits and human images with video diffusion, controllable motion and camera conditions. - Diffusion-based video editing (120 papers)
Edits video consistently with diffusion feature correspondence, reward tuning and frequency-aware factorization. - Talking face generation (155 papers)
Generates audio-driven talking faces from single images with landmark priors and motion diffusion. - Video-to-audio synthesis (117 papers)
Generates synchronized soundtracks from video with latent diffusion and text conditioning. - Text-to-motion diffusion (261 papers)
Synthesize human motion from text using diffusion models.
Most cited and most cited since 2024:
- One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing (CVPR 2021 · 445 citations)
- Hierarchical Cross-Modal Talking Face Generation With Dynamic Pixel-Wise Loss (CVPR 2019 · 440 citations)
- MagicAnimate: Temporally Consistent Human Image Animation using Diffusion Model (CVPR 2024 · 143 citations)
- Video-P2P: Video Editing with Cross-attention Control (CVPR 2024 · 123 citations)
Related topics in Video
- Video segmentation and tracking (2,850)
- Video-language understanding (1,681)
- Action recognition (1,325)
