Objective: train a simple object detection model to identify instances of products in a given image. This notebook will attempt to locate and classify occurrences of different products in supermarket shelves using a subset of the Grozi-3.2K dataset
- Special care for the directory where the repository of the algorithm used is cloned
dir_to_project >> parameter to complete with the directory of the project
dir_to_repo >> parameter to complete with the directory of the repo cloned
project -- grozi_coco -- images -- train
| | |----- validation
| |
| |-------- labels -- train
| |----- validation
yolov5
Notebook created in Google Colab which can be here:
-
Maindola, G., 2022. Introduction to YOLOv5 Object Detection with Tutorial - MLK - Machine Learning Knowledge. [online] Machinelearningknowledge.ai. Available at: https://machinelearningknowledge.ai/introduction-to-yolov5-object-detection-with-tutorial/ [Accessed 24 March 2022].
-
Gandhi, R., 2018. R-CNN, Fast R-CNN, Faster R-CNN, YOLO — Object Detection Algorithms. [online] Medium. Available at: https://towardsdatascience.com/r-cnn-fast-r-cnn-faster-r-cnn-yolo-object-detection-algorithms-36d53571365e [Accessed 24 March 2022].
-
Gur Arie, L., 2022. The practical guide for Object Detection with YOLOv5 algorithm. [online] Medium. Available at: https://towardsdatascience.com/the-practical-guide-for-object-detection-with-yolov5-algorithm-74c04aac4843 [Accessed 24 March 2022].
-
Cochard, D., 2021. YOLOv5 : The Latest Model for Object Detection. [online] Medium. Available at: https://medium.com/axinc-ai/yolov5-the-latest-model-for-object-detection-b13320ec516b [Accessed 29 March 2021].
-
Yohanandan, S., 2020. mAP (mean Average Precision) might confuse you!. [online] Medium. Available at: https://towardsdatascience.com/map-mean-average-precision-might-confuse-you-5956f1bfa9e2 [Accessed 24 March 2022].
-
GitHub. 2021. GitHub - ultralytics/yolov5: YOLOv5 🚀 in PyTorch > ONNX > CoreML > TFLite. [online] Available at: https://github.com/ultralytics/yolov5 [Accessed 30 March 2022].