C3 in yolov5, CSP Bottleneck with 3 convolutions Args: in_channels (int): Number of input channels out_channels (int): Numbe

unified ELAN structure. It supports ['basic', 'repncsp'] ELAN structure. This implementation is based on https://github.com/WongKinYiu/YO

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

This implementation is based on https://github.com/WongKinYiu/YOLO/blob/main/yolo/model/module.py.

Based on timm implementation.

Computes and stores the average and current value

A convolutional block that combines two convolution layers (kernel and point-wise conv). This implementation is based on https://github.com/l

SPPELAN module cpmprising multiple pooling and convolution layers.

Scale Norm. Args: dim (int): The dimension of the scale vector. eps (float, optional): The minimum value in clamp. Defaults to 1e

Based on https://github.com/WongKinYiu/YOLO/blob/main/yolo/model/module.py A module that combines average pooling and convolution op

Based on https://github.com/WongKinYiu/YOLO/blob/b96c8eaec16cfcabbf79947d98d2c575f0a114ad/yolo/model/module.py. A module that combine

Module that generates anchors for a set of feature maps and image sizes. The module support computing anchors at multiple sizes and aspe

Computes and stores the average and current value

A single detection head.

r""" DistributedEvalSampler is different from DistributedSampler. It does NOT add extra samples to make it evenly divisible. DistributedEv

Focus width and height information into channel space.

This layers applies global pooling over a 4D or 5D input tensor Args: pool_type (Optional[str]): Pooling type. It can be mean, rms,

Computes and stores the average and current value

ONNX Model wrapper class for inferencing.

Implementation of pooling for PoolFormer