Features

• AI Depth Estimation — Supports 20+ models (DPT, MiDaS, Depth Anything v2, etc.) with CUDA acceleration  
• Batch Processing — Smart VRAM-aware queueing for images & video
• 3D Output Modes — Half-SBS, Full-SBS, VR180, Interlaced, Anaglyph, Floating Window
• Frame Interpolation — Smooth motion with ONNX-based RIFE (2x–8x)
•  Super Resolution — Real-ESRGAN upscaling (x4 from 1080p → 4K)
• Parallax Tuning — Foreground, midground, and background depth controls
• Smart Mask Effects — Feathering and ghost suppression
• Audio Tools — Attach AAC/MP3/WAV using FFmpeg (GUI-based)
• Live Feedback — Real-time FPS, ETA, cancel/pause/resume
• Preview Modes — Anaglyph, heatmap, interlaced, SBS previews
• Built-in 3D Video Player — Supports scrubbing, VR, SBS, interlaced
• Export Ready — Outputs for Quest, YouTube 3D, MP4/MKV/AVI with GPU encoding

VisionDepth3D is completely free to use.

If you find the program helpful and want to support future development, donations are greatly appreciated!  

They go directly toward:

• New hardware for testing
• Expanding depth model support
• Continued updates & features

Thanks for supporting the project! 💙