A denoising autoencoder + adversarial losses and attention mechanisms for face swapping.

A comprehensive list of recources (papers, repositories etc.) about face restoration methods.

Suite for Windows with Real-ESRGAN, RealESRNet, RealESRAnime, BSRGAN , IRCNN, GFPGAN & RIFE. Upscaling, face restoration, frame interpolation, denoising, batch processing & GPU acceleration in one tool.

Face enhancer‏ - Denoising Auto Encoder by Tensorflow and Keras and skimage

Face recognition pipeline based on Facenet and MTCNN including image preprocessing (denoise, dehazing,...) with image augmentation techniques

This is the official repository of "eDifFIQA: Towards Efficient Face Image Quality Assessment based on Denoising Diffusion Probabilistic Models" accepted in IEEE TBIOM (Transactions on Biometrics, Behavior, and Identity Science).

Official repository of the paper "DifFIQA: Face Image Quality Assessment Using Denoising Diffusion Probabilistic Models" in proceedings of IEEE International Joint Conference on Biometrics (IJCB) 2023.

[IEEE FG'23] T2V-DDPM: Thermal to Visible Face Translation using Denoising Diffusion Probabilistic Models

[WACV '23] AT-DDPM: Restoring Faces degraded by Atmospheric Turbulence using Denoising Diffusion Probabilistic Models

AUTOMATIC1111 UI custom script for img2img around face with different "Denoising Strength" settings

Official implementation of the paper "Face Morphing Attack Detection with Denoising Diffusion Probabilistic Models"

Group project for master class "Context Aware Security Analysis for Computer Vision"

A Denoising diffusion probabilistic model paper implementation trained on an anime faces dataset

Specific Immunity Denoise Based Adversarial Defense for Face Recognition

An autoencoder is a type of artificial neural network used to learn efficient data codings in an unsupervised manner.The aim of an autoencoder is to learn a representation (encoding) for a set of data, typically for dimensionality reduction, by training the network to ignore signal “noise”. Along with the reduction side, a reconstructing side is learnt, where the autoencoder tries to generate from the reduced encoding a representation as close as possible to its original input, hence its name. Several variants exist to the basic model, with the aim of forcing the learned representations of the input to assume useful properties. Examples are the regularized autoencoders (Sparse, Denoising and Contractive autoencoders), proven effective in learning representations for subsequent classification tasks,and Variational autoencoders, with their recent applications as generative models. Autoencoders are effectively used for solving many applied problems, from face recognition to acquiring the semantic meaning of words

Use a Denoising Diffusion model to swap faces

We present a comprehensive pipeline for processing Kannada audio, leveraging state-of-the-art pre-trained models from Hugging Face. The pipeline is designed to handle multiple stages of audio processing, including denoising, speech segmentation, automatic speech recognition (ASR), punctuation, transliteration, translation, and grammar checking.

Proof of concept using Principal Component Analysis to denoise images, concept to be extended to encryption through data loss manipulation

Custom implementation of the denoising task to generate dog faces 🐶 (experiment)

Fully connected denoising autoencoder built from scratch using only NumPy and Pure Math. Backpropagation implemented manually. Trained on real face photos from the LFW dataset.

I have worked on denoising the images from Olivetti Faces Dataset by using Convolution Autoencoders

Audio denoising is a common problem faced in various applications, such as speech recognition, audio recording, and telecommunications. This project focuses on implementing a denoising technique using Python and GNU Radio Companion, an open-source software development toolkit for building software-defined radios.

In 2012 I did a research stay at CIMAT. My research focus was Probabilistic Graphical Models, and sampling methods, specially Markov Chain Monte Carlo (MCMC) techniques. A did some projects like a face tracking program using particle filters and Boltzmann Machines for image denoising. I had a great time doing this. I hope you like it too.

Face Generation with Denoising Diffusion Model

Regenerating Patched or Blacked out parts of human faces using GAN and Denoising Autoencoder.

Using Python, Tensorflow and Keras. Denoising cat face images from noisy inputs. Conv2D, BatchNormalization, MaxPooling2D, UpSampling2D,

Convolutional AutoEncoder-based networks to denoise images of obfuscated human faces

RunPod AI Video Upscale & HDR — Serverless four-stage GPU pipeline: denoise, upscale, face restore, HDR tone mapping

This project focuses on denoising animal face images using a U-Net architecture. The dataset is processed and augmented with Gaussian noise to create synthetic noisy images for training the model. The project includes functionalities for denoising along with instructions to extend the model for these tasks.

This repository contains a colab Notebook implementing Denoising Diffusion Probabilistic Models (DDPM) using Hugging Face libraries. DDPM is a generative model used for image synthesis, leveraging a diffusion process to generate high-quality images.