天涯古巷
天涯古巷 is an individual/solo developer and open-source creator, not a registered company or organization. The creator focuses on open-source contributions, technical innovation and solving complex problems, with an emphasis on AI, computer vision and large-model applications; the flagship projects are local tools for extracting, removing, generating and translating video subtitles.
At a Glance
- Individual video creators and editors
- Researchers, developers and technical users needing offline computer-vision tools
- Users processing subtitles or watermarks on Windows, macOS or Linux
- Organizations or users requiring local/private media processing
AI Tools by 天涯古巷
(1)Video Subtitle Remover
Open Source AI Subtitle Remover
Discussions
No discussions yet
Be the first to start a discussion about 天涯古巷
Latest News
Video Subtitle Remover v1.4.0 released with macOS ARM64 and Windows packages; release notes include Python 3.13-era builds and current Docker variants.
Video Subtitle Extractor v2.2.0 released, including Python 3.13 and PaddleOCRv5 updates and downloadable macOS ARM64/Windows packages.
Video Subtitle Remover added a subtitle-extraction feature and documentation link connecting users to video-subtitle-extractor.
Video Subtitle Remover improved subtitle detection and added multilingual translation support, while fixing LAMA-mode and video-processing stability issues.
Products & Services
A local GUI/deep-learning tool that detects hard-coded subtitles in video frames, performs OCR and generates SRT files. The project describes fast, automatic and precise extraction modes, with Windows, macOS and Linux-oriented deployment paths.
An AI-based local application for removing hard-coded subtitles from videos and text-like watermarks from videos or images while preserving output resolution. It offers GUI and CLI usage, downloadable releases and Docker images.
A local GUI/CLI tool that transcribes video audio into subtitles and generates SRT files using Transformer/Whisper-style models, with base, medium and large model configurations.
A local SRT subtitle translation utility: it accepts an SRT file and generates a translated SRT file without requiring a third-party API, using deep-learning translation models.
Market Position
A privacy-oriented, local and open-source alternative to cloud subtitle APIs and commercial video editors: it combines subtitle OCR/extraction, generative subtitle creation, SRT translation and AI inpainting-based subtitle removal in a small developer-maintained portfolio. Its differentiators are no required third-party API, source availability, CLI/GUI access, GPU/CPU options and Docker support; compared with hosted tools, setup and hardware/model requirements are higher.
Leadership
Founders
Yao Fang (天涯古巷 / YaoFANGUK)
The GitHub profile identifies the creator as 天涯古巷 / YaoFANGUK and lists Nottingham, UK. The official homepage describes interests in AI, images and large models and lists research coauthorship on NeurIPS 2023, ICCV 2023 and ACM Multimedia 2024 papers in computer vision and sign-language retrieval.
Executive Team
Yao Fang (天涯古巷 / YaoFANGUK)
Independent developer and project maintainer
Maintains the YaoFANGUK GitHub portfolio and the four subtitle-focused projects; the official homepage also identifies research coauthorship in computer vision and sign-language retrieval.
Founding Story
The official site presents this as a personal portfolio rather than a company. The stated motivation is to pursue open-source contribution, technical innovation and complex-problem solving; the project family applies local deep-learning models to practical video-subtitle workflows without requiring third-party APIs.
Business Model
Revenue Model
The projects are distributed as free, open-source software under Apache License 2.0. The GitHub profile also exposes a Sponsor link; no subscription, API-usage or commercial licensing pricing is stated on the reviewed official pages.
Target Markets
- Individual video creators and editors
- Researchers, developers and technical users needing offline computer-vision tools
- Users processing subtitles or watermarks on Windows, macOS or Linux
- Organizations or users requiring local/private media processing
- Extracting burned-in subtitles from video into editable SRT files
- Removing hard-coded subtitles from videos
- Removing text watermarks from images and video
- Generating subtitles from spoken video audio locally
- Translating existing SRT subtitle files locally
- Batch and offline media processing for users who do not want to send media to third-party APIs