bjfrbjx / stream-wav2lip
优化wav2lip的执行步骤，将头脸分离、嘴型替换、回补背景三个步骤分离，添加gfpgan强化面部功能，实现提前解帧，流式循环处理，对接obs
☆45Updated 5 months ago
Related projects ⓘ
Alternatives and complementary repositories for stream-wav2lip
- This is a project about talking faces. We use 576X576 sized facial images for training, which can generate 2k, 4k, 6k, and 8k digital hum…☆49Updated 8 months ago
- Just a suturing monster project.☆37Updated last year
- Considering the original Wav2Lip was trained on LSR2 and didn't have good performance on Chinese. I preprocessed CMLR Dataset and would t…☆59Updated last year
- ☆37Updated 10 months ago
- Full version of wav2lip-onnx including face alignment and face enhancement and more...☆70Updated last month
- ☆189Updated 10 months ago
- 基于MuseTalk的数字人代码。☆19Updated 2 months ago
- wav2lip训练数据预处理综合工具☆36Updated last year
- ☆67Updated last month
- python库，实现推送实时rtmp音视频流☆97Updated 7 months ago
- ☆46Updated last year
- ☆32Updated last year
- ☆30Updated 10 months ago
- 通过此代码可以免训练模型并通过轻量级服务器定制数字人形象☆99Updated 7 months ago
- Simple and fast wav2lip using new 256x256 resolution trained onnx-converted model for inference. Easy installation☆27Updated last month
- ☆193Updated last year
- 洛曦 数字人视频播放器，带HTTP API，使用gradio api对接Easy-Wav2Lip、Sadtalker、GeneFacePlusPlus、MuseTalk，也可以用于播放本地视频☆145Updated last month
- 这是一个在wav2lip，使用wav2lip、gfpgan、yolov5等模型用RT加速的超快推理！经测试在2070显卡上可达到0.03秒每帧实现实时推理。☆26Updated last year
- ☆31Updated last year
- An optimized pipeline for DINet reducing inference latency for up to 60% 🚀. Kudos for the authors of the original repo for this amazing …☆104Updated last year
- ☆65Updated 11 months ago
- ☆53Updated last year
- The API server version of the SadTalker project. Runs in Docker, 10 times faster than the original!☆120Updated last year
- ☆50Updated last year
- 基于wav2lip进行虚拟数字人训练，唇形驱动，包括数据处理流程等，模型包括96x96，192x192，192x288，288x288。☆14Updated 6 months ago
- JoyHallo: Digital human model for Mandarin☆338Updated this week