zentrocdot
banner
zentrocdot.bsky.social
zentrocdot
@zentrocdot.bsky.social
I like photography and graphic art. I am interested in AI, Crypto Coin and NFT technology. I work with Linux, use Open Source software and program myself.

Make a donation if you would like to support my work:
https://nowpayments.io/donation/zentrocdot
Pinned
In times like these, it's better to stay at home as an alien.
Last continued YOLO darknet training. I increased the number of training images. I removed duplicates. I also used better settings. Better mAP with less iterations.

#AI #YOLO #darknet #COCO #model #training #objectdetection #boundingbox #annotation #Python #Bash #iterations
December 1, 2025 at 7:03 AM
November 29, 2025 at 8:40 AM
November 29, 2025 at 8:39 AM
My object detector can be found at Copus. All resources for fast testing given.

www.copus.io/work/e105b86...

#YOLO #darknet #Python #objectdetection #copus
Python Script for the Object Detection Using YOLO
Python script for the object detection using darknet for YOLO using OpenCV
www.copus.io
November 29, 2025 at 8:20 AM
I perform a new training with modified settings. I am interested in the mAP curve to learn how I can use this curve info. Right is the calculation rule. TP is true positive and FP is false positive. The val images are used. Small set of val images makes the curve not smooth.

#YOLO #darknet #mAP
November 28, 2025 at 9:09 AM
Trainings results with continuation of training. I should have used darknet from the start.

#AI #YOLO #darknet #COCO #model #training #objectdetection #boundingbox #annotation #Python #Bash #iterations
November 27, 2025 at 1:58 PM
Influence of the training iterations on the detection results.

#AI #YOLO #darknet #COCO #model #training #objectdetection #boundingbox #annotation #Python #Bash #iterations
November 27, 2025 at 1:54 PM
I am conducting some new trainings sessions to better understand certain influencing factors. Test of the model is done with my new detector. Now I go up to 10000 iterations to see, what happens.

#AI #YOLO #darknet #COCO #model #training #objectdetection #boundingbox #annotation #Python #Bash
November 27, 2025 at 10:09 AM
I can now detect cats using my own cat models. For the detection I wrote a small Python detector on base of OpenCV using DNN as neural network. Now I can improve my model. NMS is influencing the result.

#AI #YOLO #COCO #model #training #detection #objectdetection #darknet #boundingbox #Python #Bash
November 27, 2025 at 6:56 AM
I have a false positive rate of round about 4.5 % in the cat recognition. It is for me not clear, why e.g. a rabbit is wrong recognized.

#AI #YOLO #COCO #model #objectdetection #darknet #boundingbox #Python #Bash
November 26, 2025 at 12:05 PM
This result has now amazed me. I didn't think that 100% recognition could be achieved or that 100% was achievable.

#AI #YOLO #COCO #model #objectdetection #darknet #boundingbox #Python #Bash
November 26, 2025 at 11:01 AM
Interim result if I continue the training, in this case with some new images.

#AI #YOLO #COCO #model #objectdetection #darknet #boundingbox #Python #Bash
November 26, 2025 at 9:59 AM
If there is a memory allocation problem using darknet, one should take a look at the used VRAM and at the setting subdivisions=4 in the cfg file. My setup from today failed with subdivisions=2. The 8 GB GPU was a problem.

#AI #YOLO #COCO #model #objectdetection #darknet #Linux #GPU #vram
November 26, 2025 at 9:55 AM
Here you can see a problem with darknet and GPU 0,1,2. I stopped the training and the memory is not free on the GPUs. If thios is the case I cannot restart. That was a pitfall yesterday, so I was looking for a problem that didn't exist.

#AI #YOLO #COCO #model #objectdetection #darknet #Linux #GPU
November 26, 2025 at 9:13 AM
I still need to check whether my return to training has been successful. The gray shows the failed attempt. The pink and red look as I expected.

#AI #YOLO #COCO #model #objectdetection #darknet #Linux #Python #Bash #loss #mAP
November 26, 2025 at 8:11 AM
These images show how good a YOLO training with darknet works. loss goes down, the mAP goes up. Recognition results are fine so far. Resuming training to improve is not working out at the moment.

#AI #YOLO #COCO #model #objectdetection #darknet #Linux #Python #Bash
November 26, 2025 at 7:42 AM
Best result using darknet in object recognition. The cat is recognized and the squirrel is no longer recognized. I am now continuing to pursue this approach. Now I need more tests.

#AI #YOLO #COCO #model #objectdetection #darknet #boundingbox #Python #Bash
November 25, 2025 at 5:13 PM