How to Detect
Could u explain how to detect with any image?
Hi, This is what I did to run it on my images: 1/First add a "configuration file" for your images in deep-darts/configs (one similar to the 2 yaml files already there) 2/Then you need annotate your images by running the annotate.py this will create the labels (you might have to put them in a specific directory: see what you get from the error message). This will create the true "labels" for your images. 3/ make sure to comment the assert line in dataloader.py (asserts the dataset is either d1 or d2) and add a line to handle your data. In predict.py also handle the case for your own config by setting your own value for the labels_path 4/Then you run the predict.py on your images
Hello @wmcnally ,
i test on my PC (NVIDIA RTX A2000 Laptop GPU computeCapability: 8.6)
i downloads models and cropped image from dataset1 i 've launched predict dataset1 for instance with dataloader(d1_03_03_2020) and i have 0% PCS.
I'd appreciate it if you could spare me some time on discord ? i live in france on GMT +1 Thank you very much four your help
Hey @Kariti, did you found a solution? Best regards Marcel