If you want the maximum accuracy, regardless of the speed, trainBODY_25B Model - Option 1. If you want to maintain the current OpenPose speed, while increasing its accuracy, go forBODY_25B Model - Option 2. If you are OK losing a bit of accuracy but simply want to fine-tune the...
Jacob-Chew/openpose_train 代码Issues0Pull Requests0Wiki统计流水线 服务 加入Gitee 与超过 1200万 开发者一起发现、参与优秀开源项目,私有仓库也完全免费 :) 免费加入 已有帐号?立即登录 master 分支(1) 管理 管理 master 克隆/下载 HTTPSSSHSVNSVN+SSH ...
Training repository for OpenPose. Contribute to CMU-Perceptual-Computing-Lab/openpose_train development by creating an account on GitHub.
I'm trying to fine tune the Openpose 25B model using my own data (annoted pictures). I formatted the annotation in JSON using the same layout as MPII data set, and then used the provided script in openpose_train/MPII folder to generate the mask, the final JSONS and finally the lmdb ....
def make_model(img, results, mask, is_train=True, reuse=False): confs = results[:, :, :, :n_pos] pafs = results[:, :, :, n_pos:] m1 = tf_repeat(mask, [1, 1, 1, n_pos]) m2 = tf_repeat(mask, [1, 1, 1, n_pos * 2]) cnn, b1_list, b2_list, net = model(...