v1.0.45
GargantuaX/gemini-watermark-removerv1.0.45Sep 22, 2026by GargantuaX
Upgrading?
AI Summary
This release introduces batch video processing capabilities to the standalone interface, allowing users to select and manage multiple videos sequentially. It also improves workflow stability by adding cancellation features and ensuring no partial results are saved.
Key Highlights
- Add videos to an active batch without replacing the current one; failed files do not stop the queue.
- Select or drop multiple videos in the standalone video interface with per-file status.
- Stop current detection/export and cancel pending videos to ensure no partial results are downloaded.
- Preserve the existing manual workflow for selecting a single video while idle.
New Features
- Batch video processing in the standalone interface.
- Cancel current detection/export and pending videos.
- Add videos to an active batch without replacing the current one.
- Inclusion of ONNX Runtime 1.26.0 license and third-party notices.
Full Release Notes
# 1.0.45 Release date: 2026-09-22. ## Changes since 1.0.44 - Stop current detection/export and cancel pending videos. No partial result is downloaded, and a new selection can be processed after cleanup completes. - Select or drop multiple videos in the standalone video interface. Videos are processed sequentially, with per-file status and automatic downloads. - Add videos to an active batch without replacing the video currently being processed. A failed file does not stop the remaining queue. - Preserve the existing manual workflow when selecting a single video while idle. - Include ONNX Runtime 1.26.0's complete license and upstream third-party notices alongside the redistributed runtime assets. The image-removal algorithm, video-removal algorithm, and SDK interfaces are unchanged from 1.0.44. The public website is a separate repository and deployment; batch-video support in this release is scoped to the open-source standalone interface. The website keeps its existing single-video workflow. Adding a website batch interface is not a prerequisite for this release. A future website SDK upgrade can adopt underlying fixes without exposing batch processing. Batch processing runs one video at a time for convenience; it does not accelerate AI inference. Processing can be CPU- and memory-intensive. Performance on 4K, long videos, large queues, and mobile devices has not been validated. Thanks to @kavyp12 for the batch-video contribution in #109.
Related Repositories
affaan-m/ECC★ 286,017
The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
react/react★ 251,631
The library for web and native user interfaces.
nodejs/node★ 122,533
Node.js JavaScript runtime ✨🐢🚀✨
Leonxlnx/taste-skill★ 100,805
Agent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
louislam/uptime-kuma★ 92,740
A fancy self-hosted monitoring tool