JacobLinCool/imbd-2024-c3-1000x-2

C3: Condition Classifier CNN

By Team 113036 Dataset Unavailable.

Summary

Our model utilizes a multi-scale convolutional neural network (CNN) to achieve a remarkable 99.7% accuracy and a cross-entropy loss of 0.062 on the validation set for condition classification.

Data

• Training Set: Augmented dataset from object IDs 1 to 11, with an augmentation factor of 1000x, totaling 44,000 items.
• Validation Set: Augmented dataset from object IDs 12 and 13, with the same augmentation factor of 1000x, totaling 8,000 items.

Training Metrics

Detailed training metrics and history of our model can be accessed here.