Analysis

[GabrielBG0]

Responsible for creating visualisations with data

• Unique scripts for each visualisation
• Final metrics

Features to be implemented

• F1 (from Torch Metrics)
• mIoU (from Torch Metrics)
• pixel accuracy #30
• mean class accuracy #31
• DICE (from Torch Metrics)
• R2 (from Torch Metrics)
• PSNR (from Torch Metrics)
• Loss Curve
• Final Metrics Graph #38
• Inference Display #37

[GabrielBG0]

class Metric():
    def __init__(*Args):
        ...
    def score(self, y_true: Tensor, y_pred: Tensor) -> float:
        ...