Module: TFSimilarity.losses

Contrastive learning specialized losses.

Classes

class Barlow: Computes the Barlow Loss between two batches of embeddings.
class CircleLoss: Computes the CircleLoss
class MetricLoss: Wraps a loss function in the Loss class.
class MultiSimilarityLoss: Computes the multi similarity loss in an online fashion.
class PNLoss: Computes the PN loss in an online fashion.
class SimCLRLoss: SimCLR Loss
class SimSiamLoss: SimSiam Loss
class SoftNearestNeighborLoss: Computes the soft nearest neighbors loss in an online fashion.
class TripletLoss: Computes the triplet loss in an online fashion.
class VicReg: VicReg Loss