If you,if I

ok,I generate the image by [this](https://github.com/hiroharu-kato/mesh_reconstruction/blob/master/mesh_reconstruction/render.py),and the camera is the same as described in this project.Thank you! @ShichenLiu

I don‘t know, I generated the sihouette as described in the project. This is what I want to ask, how to generate an image with only one sihouette？

I met the same problem and installed the version of GPU.However,it is still not work.Any one can help me?

Thank you for replying.I did what you said,but it didn't work at all.Do you still have other ways?Or can you send me you code? ------------------ 原始邮件 ------------------ 发件人: "gopikrishnachaganti"; 发送时间:...

Did anyone had unzip all the files?I want to know which file contains the car dataset?Or somebody can send me the dara by email.Appreciate your help.

somebody can tell me where is the car dataset.Downloading all the files is too big for me,and I only need dataset for car.Or anyone can share the dataset for me?Appreciate...

> > > 我运行convert_weight.py > > > org_weights_num = len(org_weights_mess) > > > cur_weights_num = len(cur_weights_mess) > > > 这两个参数总是不相等 > > > 在config.py中修改,改为自己训练的模型 > > > #__C.YOLO.ORIGINAL_WEIGHT = "./checkpoint/yolov3_coco.ckpt"...

> 我运行convert_weight.py > org_weights_num = len(org_weights_mess) > cur_weights_num = len(cur_weights_mess) > 这两个参数总是不相等 > 在config.py中修改,改为自己训练的模型 > #__C.YOLO.ORIGINAL_WEIGHT = "./checkpoint/yolov3_coco.ckpt" > __C.YOLO.ORIGINAL_WEIGHT = "./checkpoint/yolov3_test_loss=8.0376.ckpt-30" > #__C.YOLO.DEMO_WEIGHT = "./checkpoint/yolov3_coco_demo.ckpt" > __C.YOLO.DEMO_WEIGHT = "./checkpoint/yolov3_test_loss=8.0376.ckpt-30"...

解决了，是因为训练的类不一样，作者最开始是80类，在这个文件前几行类别就是80，如果是按照作者的voc数据来，就只有20类，改成20就可以了。用自己的数据集就改成相应的类数

demo_image.py 第23行num_classes = 80改成20，因为voc只有20类