𝑪𝑯𝑬𝑺𝑺 𝑷𝑰𝑬𝑪𝑬𝑺 𝑫𝑬𝑻𝑬𝑪𝑻𝑰𝑶𝑵 𝒀𝑶𝑳𝑶𝑽5

𝑶𝒗𝒆𝒓𝒗𝒊𝒆𝒘 : The goal of this project is able to identify chess pieces on a board or in a video using yolov5.

📁 𝑫𝒂𝒕𝒂𝒔𝒆𝒕 𝑼𝒔𝒆𝒅 : https://universe.roboflow.com/projects-ehrjz/chess-detection-i0woo

The dataset consists of 13 classes:

• 'bishop'
• 'black-bishop'
• 'black-king'
• 'black-knight'
• 'black-pawn'
• 'black-queen'
• 'black-rook'
• 'white-bishop'
• 'white-king'
• 'white-knight'
• 'white-pawn'
• 'white-queen'
• 'white-rook

𝑫𝒂𝒕𝒂 𝑷𝒓𝒆𝒑𝒂𝒓𝒂𝒕𝒊𝒐𝒏:

• Create a bounding boxes with the help of label-img And makesense.ai website according to YoloV5.
• Prepare folder structure that can be accept by YoloV5.

  𝑺𝒕𝒆𝒑𝒔 𝒕𝒐 𝒖𝒔𝒆 𝒀𝒐𝒍𝒐𝒗5:

• Cloning the YoloV5 file from official repository.
• Changing the directory of yolov5
• Installing the dependencies
• Download all versions pre-trained weights.

𝑺𝒕𝒆𝒑𝒔 𝑩𝒆𝒇𝒐𝒓𝒆 𝑻𝒓𝒂𝒊𝒏𝒊𝒏𝒈 𝑪𝒖𝒔𝒕𝒐𝒎 𝑫𝒂𝒕𝒂𝒔𝒆𝒕:

• Go to yolov5/data/.
• Open data.yaml
• Edit the following inside it:

1. Training and Validation file path
2. Number of classes and Class names.

𝑻𝒓𝒂𝒊𝒏𝒊𝒏𝒈 𝒀𝑶𝑳𝑶𝑽5 𝑴𝒐𝒅𝒆𝒍:

• Set images size 640 with batch of 8.
• Total 606 images for training and 58 images for validation present in 13 classes.
• Train model around 600 epochs .Stopping training early as no improvement observed in last 100 epochs. Best results observed at epoch 201, best model saved as best.pt.
• Visualise the training metrics with the help of tensorboard.

  𝑻𝒆𝒔𝒕𝒊𝒏𝒈 𝑰𝒎𝒂𝒈𝒆𝒔 𝑼𝒔𝒊𝒏𝒈 𝑻𝒆𝒔𝒕 𝑫𝒂𝒕𝒂:

  

𝑻𝒆𝒔𝒕𝒊𝒏𝒈 𝑽𝒊𝒅𝒆𝒐 𝑫𝒆𝒎𝒐:

chess.1.vd.mp4