Cheng Rui
Cheng Rui
if I design a 3D model myself, do I need to follow the topology of the face landmark of mediapipe?
opened the Kafuka service and replaced the so file according to the readme, but an error occurred during the operation. I referred to the previous issue, but it was not...
If you use Blendshape to calculate the face, can mediapipe not calculate multiple groups of Blendshape
nice job!I have a few questions, hope to get your reply q1.Will the training part and the CUDA accelerated rendering version be announced soon? q2.For example, how to calculate 52blendshapes...
非常感谢分享这么好的项目供给学习,由于我不知道这个PR是否会merge,所以我基于自己的使用体验,仅做如下修改: 1. 适配了Tag2Text的项目,有BUG可能是因为transfomers的版本问题,因为你们更新速度比较快。这里我提供下一种问题:”按照Tag2Text作者提供的4.15.0是完全可以推理正常的,如果安装4.15.0后导致别的比如grounded的transformer兼容问题,缓存下载后可以再切换回自己的高版本就可以了“,目前我的工作目录是兼容的。 2. huggfingface SD模型有些大了,这样调用时候容易断掉,看Issue很多同学不知道缓存在哪里并且系统的目录累积大模型占据空间并不好,所以加了接口的cache_dir指定安装,还有autoLabel脚本中的gpt是否启动; 再次感谢你们团队的工作,CVPR的论文我正在精读中,期待你们release grounded with text的训练部分,并且基于工程部分的应用和算法我也有许多想法和想做的
简单脚本供参考 可以不用merge~ 1.downloads tag2text weights 2.pip install transforms==4.15.0,run
https://github.com/positive666/Prompt-Can-Anything/assets/28972473/c9cc64af-939d-480f-a684-08d8db34b25f