Video generation
Text/image/video-to-video, motion control, long-form consistency and model pipelines
- discussionCan OpenTrackIO camera/lens metadata condition camera-controlled video generators?
- discussionDo video generation models respect or declare colour space, transfer function and range?
- discussionDo diffusion inpainters beat propagation methods for clean plates on moving cameras?
- discussionDoes pose-driven video generation (Wan-Animate, MimicMotion, UniAnimate) preserve performance nuance compared with capture-to-rig animation?
- discussionHow can a trained character LoRA be combined with performance-transfer models like Wan-Animate?
- discussionHow should pipelines record license chains when weights derive from other weights with differing terms?
- discussionHow much visual and motion quality do offloading and 4-bit/FP8 quantization cost for large open video models?
- discussionCan trajectory control and camera re-rendering be composed to build multi-angle coverage of one location and action?
- discussionDo streaming/autoregressive video models keep character and location identity over minutes better than chaining first/last-frame segments?
- discussionHow should a storyboard panel be converted into an executable shot specification an agent can pass to a video model or a 3D scene?
- discussionHow reliably do current video generators follow named camera moves and lens language in text prompts?
- sourceRecreate Our 1-Minute Seedance 4K Film: Full Breakdown With All Prompts
- sourceHiggsfield community projects gallery
- sourceMake Viral AI Short Films With This Exact Workflow (pirate short tutorial)
- sourceHow to Turn Your Own Footage Into Cinematic VFX with Seedance 4K
- sourceAI vs VFX: Can Seedance 2.5 beat VFX? (full workflow)
- sourceAI Short Film: Full Prompt Library (8-scene, 8-style animated short, Seedance 2.0)
- sourceWan-AI/Wan2.2-Animate-14B (model card)
- sourceCogVideo / CogVideoX (Zhipu AI)
- sourceTechNode: ByteDance's Seedance 2.0 hits Cannes with 95-minute AI film Hell Grind
- sourceSeedance 2.5 vs Seedance 2.0 (Higgsfield comparison)
- sourceSeedance 2.5: Complete Prompting Guide (Full Prompt Library)
- sourceGenerating with Seedance 2.0: Full Tutorial
- sourceHiggsfield Original Series: The First Complete AI Streaming Platform
- sourceDouble Shots produces music videos for pop idol Ivy with Cinema Studio
- sourceHiggsfield blog index (guides, original productions, MCP docs)
- sourceCineD: Hell Grind - the 95-minute AI feature Cannes 2026 says it never screened
- sourceWan2.1 (open video generation suite)
- sourceStability generative-models (Stable Video Diffusion, SV3D, SV4D)
- sourceSkyReels-V2 (diffusion-forcing long video)
- sourceMAGI-1 (chunk-wise autoregressive video generation)
- sourceWan: Open and Advanced Large-Scale Video Generative Models (technical report)
- sourceVideo Diffusion Models: A Survey (Melnik et al.)
- sourceLongLive (real-time interactive long video generation)
- sourceLTX-Video (v0.9.x line, superseded by LTX-2)
- sourceDiffusers guide: video generation basics and memory reduction
- sourceOpen-Sora 2.0 (11B open video model with training code)
- sourceMochi 1 (Genmo)
- sourceAnimateDiff (motion modules for image models)
- sourceSelf Forcing (autoregressive streaming video diffusion)
- sourceComfyUI docs: Wan 2.2 native video workflows
- sourceLlama Learns to Direct: DirectorLLM for Human-Centric Video Generation
- sourceATI: Any Trajectory Instruction
- sourceAutomated Movie Generation via Multi-Agent CoT Planning (MovieAgent)
- sourceOne of One: Maradona Official brand world from concept to sold out
- sourceVertexCGI builds a viral NBA Finals campaign in a five-day sprint
- sourceHiggsfield Supercomputer Guide: make a complete AI video without switching tools
- sourceSinecera runs hybrid production on Higgsfield (Lacoste x SuperStep, Nesine)
- sourceHiggsfield Creator Hub: customer case-study directory
- sourceBig Purple: Turkiye Wealth Fund anniversary film in 10-12 days with Canvas