A scalar meter uses a deque to track a series of scaler values with a given window size. It supports calculating the median and average value

Abstract base class for all deepfake datasets.

Fuses the information from the Fast pathway to the Slow pathway. Given the tensors from Slow pathway and Fast pathway, fuse information from

Flat tensor

Residual block.

A named tuple describing a ResNet block.

Dual CMA attention Layer

SE block

Component model mask

DFL facial mask

Extended mask Based on components mask. Attempts to extend the eyebrow points up the forehead

Basic face hull mask

Dual CMA attention Layer

Apply gaussian smoothing on a 1d, 2d or 3d tensor. Filtering is performed seperately for each channel in the input using a depthwise c

Apply gaussian smoothing on a 1d, 2d or 3d tensor. Filtering is performed seperately for each channel in the input using a depthwise conv

Graph Reasoning Module for information aggregation.

Reconstruction Guided Attention.

Wraps a callable in an :class:`nn.Module` without registering it.

Wraps a callable in an :class:`nn.Module` without registering it.

Convolution block without no-linear activation layer

Builds Non-local Neural Networks as a generic family of building blocks for capturing long-range dependencies. Non-local Network computes

End-to-End Reconstruction-Classification Learning for Face Forgery Detection

ResNe(X)t 3D stem module. Performs spatiotemporal Convolution, BN, and Relu following by a spatiotemporal pooling.

VAE_Decoder: Decode noise to image

VAE_Encoder: Encode image into std and logvar

Xception optimized for the ImageNet dataset, as specified in https://arxiv.org/pdf/1610.02357.pdf

Component model mask

DFL facial mask

Extended mask Based on components mask. Attempts to extend the eyebrow points up the forehead

Basic face hull mask

Computes the AM-Softmax loss with cos or arc margin

Computes the AM-Softmax loss with cos or arc margin

Measure the AVA train, val, and test stats.

All backbones for detectors should subclass this class.