Project that detects the brand of a car, between 1 and 49 brands, that appears in a photograph, with a success rate of more than 70% (using a test file that has not been involved in the training as a valid or training file, "unseen data") and can be implemented on a personal computer
machine-learning brand deep-learning tensorflow keras callback classification imagenet image-classification attention cars transfer-learning pretrained-models inceptionv3 cnn-for-visual-recognition early-stopping stanford-dataset
-
Updated
Mar 30, 2024 - Python