Merged
Conversation
updated french json to include missing keys
updated json to add missing keys
updated json to add missing keys
update json to add missing keys
# 🛠️ VisionDepth3D v3.2.X – Performance & Stability Update ## Changelog: ### Fixed UI Freezing During Folder Processing - Resolved a bug where the GUI would freeze when processing a folder of videos for depth generation. - Properly offloaded batch size parsing and folder selection to the main thread, with core logic in a background thread. ### Removed Redundant Batch Logic in Folder Processing - Cleaned up double handling of `batch_size` logic in `process_video_folder()` and `process_videos_in_folder()`. - Now batch size is handled only once, before launching the background thread, improving clarity and reliability. ### Refactored ONNX Model Loader - Rebuilt the `load_onnx_model()` function to support both rank-4 and rank-5 ONNX models. - Adds automatic padding for 32-frame models (e.g. Marigold) and returns predictions as `torch.tensor` objects. - Includes robust shape detection, better error handling, and future support for diverse ONNX pipelines.
🎬 Integrated PySceneDetect into FrameTools GUI for Scene-Based Splitting - Added support for automatic scene detection using PySceneDetect. - Clips can now be segmented into manageable scenes directly from the FrameTools tab. - Enables smarter depth processing by allowing fine-tuned 3D rendering per scene. - Added user-adjustable threshold slider (NOTE: fixed logic to reflect correct cut behavior).
FrameTools Tab: Improved Frame Extraction Workflow - Added directory selector to choose custom output location for extracted frames. - Switched to FFmpeg for frame extraction, significantly improving speed and reliability.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
VisionDepth3D v3.2.5 – Changelog
Note: This version focuses on backend stability, batch processing reliability, and model loading improvements, with early groundwork for advanced pipeline modularity.
GUI & Workflow Improvements
🧵 Threaded Folder Processing (No More Freezing)
batch_sizeand folder paths now handled on the main thread, with actual processing delegated to a background thread.📁 Cleaned Up Redundant Batch Logic
batch_sizebetweenprocess_video_folder()andprocess_videos_in_folder().batch_sizeis now resolved once before launching threaded tasks.ONNX Model Pipeline Upgrade
⚡ Refactored
load_onnx_model()Systemload_onnx_model()supports:torch.tensor, making it compatible with existing render logic.FrameTools Tab Enhancements
🎞️ Frame Extraction Overhaul
Experimental Integration: PySceneDetect
🎬 Scene-Aware Splitting (WIP)
🌐 Language File Updates
languages/fr.jsonand other localized files to matchen.jsonkey structure.