sigma

with tf.control_dependencies([rpn_labels]): with tf.variable_scope('sample_RCNN_minibatch'): rois, labels, bbox_targets = \ tf.py_func(proposal_target_layer, [rois, gtboxes_batch], [tf.float32, tf.float32, tf.float32]) rois = tf.reshape(rois, [-1, 4]) labels = tf.to_int32(labels) labels = tf.reshape(labels, [-1]) bbox_targets = tf.reshape(bbox_targets,...

In best_model_config.yaml num_feature_levels: 1 In config.py config.DECODER.use_feat_level = [0, 1, 2] I am confused about that if use_feat_level = [0, 1, 2], should the num_feature_levels be equal to 3 ?...

In line 326 of the main.cpp file, int detCount = output[0]; can you explain why output[0] is detcount?thanks you.

请问使用Relu layer, Scale layer 和 ElementWise替换prelu需要重新训练吗？

2

请问使用Relu layer, Scale layer 和 ElementWise替换prelu需要重新训练吗？还是直接修改prototxt中对应的层就可以了？谢谢

Question of Equation 11

2

Thanks for your great work,There are two questions that I don't understand. I want to ask you for advice (1)As Mentioned in the paper [X‘; X]denotes the matrix concatenation, I...

In line 68 of the runDet.cpp file, int num_det = static_cast(outputData[0]); can you explain why outputData[0] is num_det? thanks you.

Will the pre-trained monocular 3D pose model on Freeman be released？

1

Thanks for your work. I would like to test the generalization of the monocular 3D pose model trained on the Freeman dataset, will the pre-trained model be released? thanks

文本位置编码的疑问

8

您好，t2v的diffusers版本中看到了对视频添加rope位置编码，但文本没找到在哪里处理的，请问文本的位置编码是全为0吗？谢谢 

请问使用多少张卡，大概训练了多久呢，谢谢

感谢您的工作，我看代码中只对视频部分做了rope位置编码，text和refer image没有看到添加位置编码，请问为什么text和refer不需要添加位置编码呢？ https://github.com/SkyworkAI/SkyReels-A2/blob/8e683d57a971ce975732b4e57638f27d394bfae3/models/transformer_a2.py#L658C11-L658C29