-
![image](https://user-images.githubusercontent.com/16919236/80744546-66b57000-8b51-11ea-9d09-30995a199bab.png)
The Class and Obj should be closed to 1. Am i right?
Why the total_loss is 0 when…
-
### Releasing v3.0.0-beta.3 🎉️ , featuring:
- **Complete rewrite of Docker setup / install process** in order to use `docker-compose`
- Update darknet fork once again using the new -dontdraw_bbox …
-
Hello,
I tested multiple pre-trained weights provided by tkDNN with "map_demo."
However, mAP of some networks are lower than expected.
(mobilenetv2ssd512, yolo3tiny, dla34_cnet, resnet101_cnet)
…
-
my system is win10 64bit, and my dataset has 5 classes, trainset has about 1000 images, 1920*1080 pixels, so I set max_batch=10000.
but when I start to train,(command: darknet.exe detector train cfg/…
-
Implement TridentNet with Yolo-head
* resnet152_trident.cfg: https://github.com/AlexeyAB/darknet/files/3292303/resnet152_trident.cfg.txt
* The **1st place** in MS COCO detection: https://papers…
-
In get_network_boxes, there is a parameter called "hier_thresh" - please could you explain what this is - varying from 0.1 to 0.9 doesn't seem to affect my detections.
I know that NMS is for trimmi…
-
Does this repo. support CSPResNeXt50-PANet-SPP? (https://github.com/WongKinYiu/CrossStagePartialNetworks/)
AlexeyABs support: https://github.com/AlexeyAB/darknet/issues/4406
My tests have found …