Changes from github community
- Fit colab demos in new env by @taoboyang in #82
- Add controlnet module by @xiaoweiw-nv in #79
- 质量评价模型合并 by @zhitianwu in #90
- Clean repo by @HuHeng in #96
- add openvino module in face_detect by @mmdzzh in #101
- vqa_demo: update readme by @HuHeng in #97
- add bmf cpu Dockerfile by @HuHeng in #98
Changes from bytedance internal
- Fix inconsistency logic in ffmpeg_encoder with ffmpeg in graph with no ffmpeg_filter
- ffmpeg_filter: error eof handling
- Fix oom bad case in generator mode
- Fix bad case in push_mode rawstream
- Add push frame mode in ffmpeg_encoder
- Add single in single out ffmpeg filter api and python binding
- Fix ffmpeg filter merge bug
- The input of the ffmpeg_filter is no longer required to necessarily contain the private AVFrame.
- Add merge ci for windows platform, fix problems with windows test case.
- Fix running problems in colab demos
- update the distro of base image used for release from centos7 to Almalinux 8
- upgrade the version of cuda in build pipeline to 12.2
New Contributors
- @zhitianwu made their first contribution in #90
- @mmdzzh made their first contribution in #101
Full Changelog: v0.0.9...v0.0.11
Contributors
HuHeng, mmdzzh, and 3 other contributors