zeyang

Data Chunks Generation

1

Hi, I find it hard to understand what is going on with the data chunks generation process in the file below. https://github.com/daveredrum/Pointnet2.ScanNet/blob/f4adc43779aaaefd6138d37183a94185d0718865/lib/dataset.py#L219-L240 1. What is line 236 meant for? 2....

Try to reproduce voc 10-1 results

3

I am trying to reproduce 10-1 results as shown in the table below. I notice a large gap of the old class mIoU between my reproduce result (38.82) and your...

Problem with First Stage Training

2

Hi, intuitively, I first download the model you provide and finetune on the few-shot dataset so as to reproduce your results after finetuning. Basically, I can reproduce your results with...

qpth Error: Cannot perform LU factorization on Q.

4

Pytorch 1.0.1. Still have this problem with RTX2080ti.

Report data leakage that causes unfair few-shot setting

1

I think there exists a data leakage problem in your code, which severely hurts the fair comparison with other methods. Take coco 10shot for example, you will first construct a...

Results un-reproducible after your revision on Nov 6, 2019

3

As I report in the previous issue, I notice that you have fixed it. However, this will lead to significant decrease of the performance due to the reduce of annotations,...

I think there exists a data leakage problem in your code, which severely hurts the fair comparison with other methods. Take coco 10shot for example, you will first construct a...

ImportError: sys.meta_path is None, Python is likely shutting down

10

 When I run this code, I encounter this error. I don't know how to fix it. Your help is highly appreciated.

Report a bug in meta training phase 2

2

There is a bug in the function filter_class_roidb(roidb, shot, imdb) in line 59 of file MetaRCNN/lib/roi_data_layer/roidb.py, resulting in actual novel shots exceeding designated shots. For instance, supposing we have collect...

Hi, I run the code with 8x TITAN Xp. The batch size is set to 8. I come across the error as shown below.