-
I also encountered the same problem with #85 using my own dataset.
Furthermore, in my case, inf didn't appear in the loss output because of the choices in hard negative mining although the inference…
-
```
STOP! If your issue is with URI rewriting, read the wiki first:
http://code.google.com/p/minify/wiki/CommonProblems#URIs_are_re-
written_incorrectly_in_CSS_output
Please remove this message from t…
-
```
STOP! If your issue is with URI rewriting, read the wiki first:
http://code.google.com/p/minify/wiki/CommonProblems#URIs_are_re-
written_incorrectly_in_CSS_output
Please remove this message from t…
-
I have a question about the performance of deconvolution applied in ms-cnn.
did you check the performance of object detection with deconvolution on KITTI benchmark test images?
I want to know how mu…
-
**Obective**: Training SSD Network to be able to detect text, fillable text fields and check boxes in a scanned document.
# TL;DR:
Trained SSD network on my dataset it MutliboxTarget function retu…
-
-
forward
loss_c[pos] = 0 # filter out pos boxes for now
IndexError: The shape of the mask [1, 8732] at index 0 does not match the shape of the indexed tensor [8732, 1] at index 0
How ca…
-
- [x] Explain Obeject Detection concept
- [ ] Explain Obeject Detection Model([Faster R-CNN](https://arxiv.org/pdf/1506.01497.pdf)) ➡️ DateSet([Deepfashion2](https://arxiv.org/pdf/1901.07973.pdf))
-…
-
Thank you for your project. It runs good on Windows. I’m working on ssd (Single Shot MultiBox Detector)now. Accuracy rate of ssd is as same as that of faster Rcnn while its speed is as quick as yolo’s…
-
I'm not sure how to repro, but I've had a few instances where after running for a while and closing the game to jump between regions, the tool is running but doesn't find any handle.
However if I o…