marcovaldo

@HawkAaron Why should we use log_softmax when training at [here](https://github.com/HawkAaron/RNN-Transducer/blob/graves2013/model.py#L77). When training the CTC model, there is no activation function such as softmax and log_softmax at [here](https://github.com/HawkAaron/RNN-Transducer/blob/graves2013/model.py#L20) Is there something...

A question about weight initialization of embedding layer.

1

I have a question about the weight intialization of embedding layer. In source code of PyTorch, the weight of embedding layer is initialized by [N(0, 1)](https://github.com/pytorch/pytorch/blob/master/torch/nn/modules/sparse.py#L105). In this code, the...

the role of mask in attention operation

2

I am reading torch implementation, your implementation and the pytorch implementation. I found that there are mask in your implementation and torch implementation, but there is no mask in pytorch...

I have trained a model which performance was good, can I finetune this pretrained model on specaugmented data to get better performance? How did you get your result, did you...

As the title says, can we finetune the pretrained model on specaugmented data? I did this experiment, but I didn't get improvement. How did you get your result, did you...

a problem about computing a person's center

5

“human_annotations”：若干长度为4的整数型数列，存储人体框的位置。其中前两个参数为人体框左上角点的坐标值，后两个参数为人体框右下角点的坐标值。 "human_annotations"：contains the positions of human bounding boxes. The first two parameters indicate the top left coordinates of the human bounding box, and the last two parameters are the lower...