🎮 Wan2.2 14B Fun Control Video - GGUF Workflow with Canny Edge Detection
Take full control of your video generation with this advanced Wan2.2 workflow featuring real-time control video guidance!
✨ Key Features:
• Control Video Integration: Use any video as motion/structure guidance
• Canny Edge Processing: Automatic edge detection for precise control
• Reference Image Support: Maintain character/style consistency
• GGUF Optimized: Memory-efficient Q8_0 quantization
• Dual-Stage Pipeline: High/Low noise models for premium quality
• 4-Step LoRA Speed: Ultra-fast generation with distilled LoRAs
🔧 Technical Specs:
• Model: Wan2.2-Fun-A14B-Control (High/Low Noise)
• Quantization: Q8_0 GGUF format
• Control Method: Canny edge detection
• Resolution: 480x480 (customizable)
• Frames: 81 frames at 16fps
• Processing: Automatic video preprocessing
💡 Use Cases:
• Dance video recreation
• Motion transfer
• Character animation
• Structure-guided generation
• Style transfer with motion control
📋 Requirements:
• Control video (MP4)
• Reference image
• ComfyUI with ControlNet aux nodes
• GGUF models from QuantStack
Transform any video into your creative vision while maintaining perfect motion control!
#Wan22 #ControlNet #GGUF #ComfyUI #VideoControl #AI
Description
FAQ
Comments (3)
do you have an settings recommendations for quality, the image you have for a demo didnt seem to use this workflow haha
I used this workflow to create the video. In fact, I simply upscaled the final result using my upscaler workflow, which you can find here: [WAN 22.5B Latent Video Upscaler and Enhancer](https://civitai.com/models/1906090/wan-22-5b-latent-video-upscaler-and-enhancer-transform-low-res-videos-into-hd-masterpieces-the-intelligent-way).
Hello, I am trying to use your workflow. Managed to make it work for the Canny model but I would like to use the depth control. How Do I do that? How do I switch the control net to use the depth model? I am quiet new to this. Thank you!