Adrian Haarbach

Use Faster-RCNN directly

7

Following up on https://github.com/pytorch/fairseq/issues/759#issuecomment-589498214, it would be great if Faster-RCNN could be used directly, so we could input images instead of pre-computed features from MS-COCO. Regarding the specific Faster-RCNN PyTorch...

This ensures OpenMP 2.5 compatibility instead of 3+ and fixes: Error C3016 'i': index variable in OpenMP 'for' statement must have signed integral type. This fixes #6.

Allow `import` statements

8

To allow `import` statements, 1. the babel parser option `sourceType` (https://babeljs.io/docs/en/options#sourcetype) needs to be configurable, which is not possible currently as it is hard-coded to `script` currently in the qx...

Table focus indicator offset

4

**Describe the bug** Sometimes there is an offset between the table cell focus indicator and the row selection. **To Reproduce** Steps to reproduce the behavior: 1. Go to https://qooxdoo.org/qxl.playground/#Table-ria 2....

Currently qooxdoo-compiler as in npm is broken after initial install ```sh sudo npm install -g @qooxdoo/compiler $ qx internal/modules/cjs/loader.js:638 throw err; ^ Error: Cannot find module 'update-notifier' at Function.Module._resolveFilename (internal/modules/cjs/loader.js:636:15)...

Load weights from original pretrained model

11

Thanks for your effort to migrate the slightly modified Faster-RCNN from Caffe to PyTorch! Your [README.md](https://github.com/airsplay/py-bottom-up-attention/blob/master/README.md) states that > The detectron2 system with **exact same model and weight** as the...