diff --git a/test/test_prototype_transforms_consistency.py b/test/test_prototype_transforms_consistency.py
index d3100bc81d5..143f1279c54 100644
--- a/test/test_prototype_transforms_consistency.py
+++ b/test/test_prototype_transforms_consistency.py
@@ -251,6 +251,8 @@ def __init__(
             ArgsKwargs(p=0),
             ArgsKwargs(p=1),
         ],
+        # Use default tolerances of `torch.testing.assert_close`
+        closeness_kwargs=dict(rtol=None, atol=None),
     ),
     ConsistencyConfig(
         prototype_transforms.RandomAdjustSharpness,
diff --git a/torchvision/prototype/transforms/functional/_color.py b/torchvision/prototype/transforms/functional/_color.py
index e46dfd74935..671f3fcf90f 100644
--- a/torchvision/prototype/transforms/functional/_color.py
+++ b/torchvision/prototype/transforms/functional/_color.py
@@ -377,17 +377,23 @@ def autocontrast_image_tensor(image: torch.Tensor) -> torch.Tensor:
         return image
 
     bound = _FT._max_value(image.dtype)
-    dtype = image.dtype if torch.is_floating_point(image) else torch.float32
+    fp = image.is_floating_point()
+    float_image = image if fp else image.to(torch.float32)
 
-    minimum = image.amin(dim=(-2, -1), keepdim=True).to(dtype)
-    maximum = image.amax(dim=(-2, -1), keepdim=True).to(dtype)
+    minimum = float_image.amin(dim=(-2, -1), keepdim=True)
+    maximum = float_image.amax(dim=(-2, -1), keepdim=True)
 
-    scale = bound / (maximum - minimum)
     eq_idxs = maximum == minimum
+    inv_scale = maximum.sub_(minimum).div_(bound)
     minimum[eq_idxs] = 0.0
-    scale[eq_idxs] = 1.0
+    inv_scale[eq_idxs] = 1.0
+
+    if fp:
+        diff = float_image.sub(minimum)
+    else:
+        diff = float_image.sub_(minimum)
 
-    return (image - minimum).mul_(scale).clamp_(0, bound).to(image.dtype)
+    return diff.div_(inv_scale).clamp_(0, bound).to(image.dtype)
 
 
 autocontrast_image_pil = _FP.autocontrast