Computes and stores the average and current value

Transformer-based Object Re-Identification

Center loss. Reference: Wen et al. A Discriminative Feature Learning Approach for Deep Face Recognition. ECCV 2016. Args: num_cl

CNN Feature Map Embedding Extract feature map from CNN, flatten, project to embedding dim.

Image to Patch Embedding with overlapping patches

Randomly selects a rectangle region in an image and erases its pixels. 'Random Erasing Data Augmentation' by Zhong et al. See https:/

Triplet loss using HARDER example mining, modified based on original triplet loss using hard example mining

r"""Implement of large margin arc distance: : Args: in_features: size of each input sample out_features: size of each

r"""Implement of large margin cosine distance: : Args: in_features: size of each input sample out_features: size of each output sa

Cosine decay with restarts. This is described in the paper https://arxiv.org/abs/1608.03983. Inspiration from https://github.com/all

Cross entropy loss with label smoothing regularizer. Reference: Szegedy et al. Rethinking the Inception Architecture for Computer Vision. CVP

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).

CNN Feature Map Embedding Extract feature map from CNN, flatten, project to embedding dim.

Image to Patch Embedding with overlapping patches

Randomly sample N identities, then for each identity, randomly sample K instances, therefore batch size is N*K. Args: - data_source (

Randomly sample N identities, then for each identity, randomly sample K instances, therefore batch size is N*K. Args: - data_source (

# Transformer-based Cross-modal Ship Re-Identification

Base class of reid dataset

Base class of image reid dataset

HOSS dataset

NLL loss with label smoothing.

Image to Patch Embedding

Image to Patch Embedding

Parameter Scheduler Base Class A scheduler base class that can be used to schedule any optimizer parameter groups. Unlike the builtin PyTorc