Detectron.pytorch

Detectron.pytorch copied to clipboard

How to train with a smaller net-input size such as (640,480)?

12

Thanks for sharing your great job. By printing the size of image, I get 768x1344 of resnet50-fpn model. However, in my case, I want to retrain this network using a...

JackHenry1992

Train on COCO Stuff + Things

3

Hello, I wish to train on the 172 classes of COCO Stuff + Things https://github.com/nightrome/cocostuff. I downloaded their 2017 training and validation COCO version annotations and added them to the...

kevinj22

How is the segmentation evaluated in mask rcnn?

1

Can anyone tell as to what metric is used for **evaluating the Segmentation masks in Mask-RCNN** and how is it implemented, because in the code, I can see that its...

vsd550

## PLEASE FOLLOW THESE INSTRUCTIONS BEFORE POSTING 1. **Read the README.md thoroughly ! README.md is not a decoration.** 2. Please search existing *open and closed* issues in case your issue...

liaozp1

Add support to python2: 1. change the way of `super()` in every class; 2. add from `__future__ import absolute_import` to import `logging` and `collections` correctly; 3. add from `__future__ import...

HongwenZhang

Update subprocess.py

1

Description: I use 3 GPUs to train the network and interrupt at some point before the final step, which means I only save the checkpoint but not config. Then, I...

Constannnnnt

fix resnext config error

1

A tiny misalignment that ought to be corrected.

w-hc

Hi, is there an implementation for non-local neural networks?

stevenwudi

the roialign in lib/model/roi_align is totally same as the caffe2, but the roialign in lib/nn/modeling/roi_xfrom/ is more similar to crop_and_resize whose sample ratio is one. So why not use roialign...

zyxunh

Error while loading custom dataset

2

Hello, I would like to run detectron on my own dataset. I just have to detect and segment a single class. In order to do that I converted my dataset...

stefat77