deepinv.metric

Metrics are generally used to evaluate the performance of a model, or as the distance function inside a loss function. Please refer to the user guide for more information.

Base class

User Guide: refer to Metrics for more information.

deepinv.loss.metric.Metric

Base class for metrics.

Full Reference Metrics

User Guide: refer to Full Reference Metrics for more information.

deepinv.loss.metric.MSE	Mean Squared Error metric.
deepinv.loss.metric.NMSE	Normalized Mean Squared Error metric.
deepinv.loss.metric.MAE	Mean Absolute Error metric.
deepinv.loss.metric.PSNR	Peak Signal-to-Noise Ratio (PSNR) metric.
deepinv.loss.metric.SNR	Compute the signal-to-noise ratio (SNR)
deepinv.loss.metric.SSIM	Structural Similarity Index (SSIM) metric using torchmetrics.
deepinv.loss.metric.L1L2	Combined L2 and L1 metric.
deepinv.loss.metric.LpNorm	\(\ell_p\) metric for \(p>0\).
deepinv.loss.metric.LPIPS	Learned Perceptual Image Patch Similarity (LPIPS) metric.
deepinv.loss.metric.SpectralAngleMapper	Spectral Angle Mapper (SAM).
deepinv.loss.metric.ERGAS	Error relative global dimensionless synthesis metric.
deepinv.loss.metric.HaarPSI	HaarPSI metric with tuned parameters.
deepinv.loss.metric.CosineSimilarity	Cosine similarity metric.

No Reference Metrics

User Guide: refer to No Reference Metrics for more information.

deepinv.loss.metric.NIQE	Natural Image Quality Evaluator (NIQE) metric.
deepinv.loss.metric.QNR	Quality with No Reference (QNR) metric for pansharpening.
deepinv.loss.metric.BlurStrength	No-reference blur strength metric for batched images.
deepinv.loss.metric.SharpnessIndex	No-reference sharpness index metric for 2D images.