[brief] Updates the unit tests to ensure the MAE metrics are tested p…

…roperly.

[detailed]

test/test_metrics.py

@@ -62,15 +62,21 @@ def test_mAP(self) -> None:
         mAP = metrics.CalculateMAP()
         self.check_almost_equal(mAP(targs, preds), functional.calculate_mAP(targs, preds))
 
-    def test_MAE(self) -> None:
+    def check_mae(self, scale: float) -> None:
         rng.seed_rngs()
 
-        pred = torch.rand((32, 32))
-        gt = torch.rand((32, 32))
-        mae = metrics.CalculateMAE()
+        pred = torch.randn((1, 3, 32, 32)) * scale
+        gt = torch.randn((1, 3, 32, 32)) * scale
+        mae = metrics.CalculateMAE(scale)
 
-        self.check_almost_equal(mae(pred, gt), functional.calculate_mae_torch(pred, gt))
+        self.check_almost_equal(
+            mae(pred, gt), functional.calculate_mae_torch(pred, gt, scale)
+        )
 
         pred = pred.numpy()  # type: ignore[assignment]
         gt = gt.numpy()  # type: ignore[assignment]
-        self.check_almost_equal(mae(pred, gt), functional.calculate_mae(pred, gt))  # type: ignore[arg-type]
+        self.check_almost_equal(mae(pred, gt), functional.calculate_mae(pred, gt, scale))  # type: ignore[arg-type]
+
+    def test_MAE(self) -> None:
+        self.check_mae(1.0)
+        self.check_mae(255.0)