I have trained two versions of Yolov7 models for road object detection. One is the main version of Yolov7 and the other one is the modified version of the main version of Yolov7. The main model has given a higher [email protected] but with lower precision on the test dataset. However, the modified version has given Higher Precision but with lower [email protected] on the test dataset. Is the modified model better than the main model in any perspective?
This is the result of the main Yolov7 model:
This is the result of the modified Yolov7 model:
Please, kindly analyze the results and tell me if the modified model is better than the main model from any point of view because of having the higher precision?