PyTorch Implementation of Vector Quantized Variational AutoEncoders.

Overview

Pytorch implementation of VQVAE.

This paper combines 2 tricks:

  1. Vector Quantization (check out this amazing blog for better understanding.)
  2. Straight-Through (It solves the problem of back-propagation through discrete latent variables, which are intractable.)

architecture

This model has a neural network encoder and decoder, and a prior just like the vanila Variational AutoEncoder(VAE). But this model also has a latent embedding space called codebook(size: K x D). Here, K is the size of latent space and D is the dimension of each embedding e.

In vanilla variational autoencoders, the output from the encoder z(x) is used to parameterize a Normal/Gaussian distribution, which is sampled from to get a latent representation z of the input x using the 'reparameterization trick'. This latent representation is then passed to the decoder. However, In VQVAEs, z(x) is used as a "key" to do nearest neighbour lookup into the embedding codebook c, and get zq(x), the closest embedding in the space. This is called Vector Quantization(VQ) operation. Then, zq(x) is passed to the decoder, which reconstructs the input x. The decoder can either parameterize p(x|z) as the mean of Normal distribution using a transposed convolution layer like in vannila VAE, or it can autoregressively generate categorical distribution over [0,255] pixel values like PixelCNN. In this project, the first approach is used.

The loss function is combined of 3 components:

  1. Regular Reconstruction loss
  2. Vector Quantization loss
  3. Commitment loss

Vector Quantization loss encourages the items in the codebook to move closer to the encoder output ||sg[ze(x) - e||^2] and Commitment loss encourages the output of the encoder to be close to embedding it picked, to commit to its codebook embedding. ||ze(x) - sg[e]]||^2 . commitment loss is multiplied with a constant beta, which is 1.0 for this project. Here, sg means "stop-gradient". Which means we don't propagate the gradients with respect to that term.

Results:

The Model is trained on MNIST and CIFAR10 datasets.

Target 👉 Reconstructed Image


👉

👉

gif

Details:

  1. Trained models for MNIST and CIFAR10 are in the Trained models directory.
  2. Hidden size of the bottleneck(z) for MNIST and CIFAR10 is 128, 256 respectively.
Owner
Vrushank Changawala
Vrushank Changawala
(CVPR2021) Kaleido-BERT: Vision-Language Pre-training on Fashion Domain

Kaleido-BERT: Vision-Language Pre-training on Fashion Domain Mingchen Zhuge*, Dehong Gao*, Deng-Ping Fan#, Linbo Jin, Ben Chen, Haoming Zhou, Minghui

250 Jan 08, 2023
PIKA: a lightweight speech processing toolkit based on Pytorch and (Py)Kaldi

PIKA: a lightweight speech processing toolkit based on Pytorch and (Py)Kaldi PIKA is a lightweight speech processing toolkit based on Pytorch and (Py)

336 Nov 25, 2022
Official Matlab Implementation for "Tiny Obstacle Discovery by Occlusion-aware Multilayer Regression", TIP 2020

Tiny Obstacle Discovery by Occlusion-aware Multilayer Regression Official Matlab Implementation for "Tiny Obstacle Discovery by Occlusion-aware Multil

Xuefeng 5 Jan 15, 2022
Built a deep neural network (DNN) that functions as an end-to-end machine translation pipeline

Built a deep neural network (DNN) that functions as an end-to-end machine translation pipeline. The pipeline accepts english text as input and returns the French translation.

Afropunk Technologist 1 Jan 24, 2022
StyleSwin: Transformer-based GAN for High-resolution Image Generation

StyleSwin This repo is the official implementation of "StyleSwin: Transformer-based GAN for High-resolution Image Generation". By Bowen Zhang, Shuyang

Microsoft 349 Dec 28, 2022
Transformer part of 12th place solution in Riiid! Answer Correctness Prediction

kaggle_riiid Transformer part of 12th place solution in Riiid! Answer Correctness Prediction. Please see here for more information. Execution You need

Sakami Kosuke 2 Apr 23, 2022
Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation

Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation [Arxiv] [Video] Evaluation code for Unrestricted Facial Geometry Reconstr

Matan Sela 242 Dec 30, 2022
Codes for realizing theories learned from Data Mining, Machine Learning, Deep Learning without using the present Python packages.

Codes-for-Algorithms Codes for realizing theories learned from Data Mining, Machine Learning, Deep Learning without using the present Python packages.

Tracy (Shengmin) Tao 1 Apr 12, 2022
The Unreasonable Effectiveness of Random Pruning: Return of the Most Naive Baseline for Sparse Training

[ICLR 2022] The Unreasonable Effectiveness of Random Pruning: Return of the Most Naive Baseline for Sparse Training The Unreasonable Effectiveness of

VITA 44 Dec 23, 2022
Code + pre-trained models for the paper Keeping Your Eye on the Ball Trajectory Attention in Video Transformers

Motionformer This is an official pytorch implementation of paper Keeping Your Eye on the Ball: Trajectory Attention in Video Transformers. In this rep

Facebook Research 192 Dec 23, 2022
Rethinking Transformer-based Set Prediction for Object Detection

Rethinking Transformer-based Set Prediction for Object Detection Here are the code for the ICCV paper. The code is adapted from Detectron2 and AdelaiD

Zhiqing Sun 62 Dec 03, 2022
Deep learned, hardware-accelerated 3D object pose estimation

Isaac ROS Pose Estimation Overview This repository provides NVIDIA GPU-accelerated packages for 3D object pose estimation. Using a deep learned pose e

NVIDIA Isaac ROS 41 Dec 18, 2022
Multi-resolution SeqMatch based long-term Place Recognition

MRS-SLAM for long-term place recognition In this work, we imply an multi-resolution sambling based visual place recognition method. This work is based

METASLAM 6 Dec 06, 2022
Human Action Controller - A human action controller running on different platforms.

Human Action Controller (HAC) Goal A human action controller running on different platforms. Fun Easy-to-use Accurate Anywhere Fun Examples Mouse Cont

27 Jul 20, 2022
Constructing interpretable quadratic accuracy predictors to serve as an objective function for an IQCQP problem that represents NAS under latency constraints and solve it with efficient algorithms.

IQNAS: Interpretable Integer Quadratic programming Neural Architecture Search Realistic use of neural networks often requires adhering to multiple con

0 Oct 24, 2021
Official PyTorch implementation of "Improving Face Recognition with Large AgeGaps by Learning to Distinguish Children" (BMVC 2021)

Inter-Prototype (BMVC 2021): Official Project Webpage This repository provides the official PyTorch implementation of the following paper: Improving F

Jungsoo Lee 16 Jun 30, 2022
Implementation of NÜWA, state of the art attention network for text to video synthesis, in Pytorch

NÜWA - Pytorch (wip) Implementation of NÜWA, state of the art attention network for text to video synthesis, in Pytorch. This repository will be popul

Phil Wang 463 Dec 28, 2022
Multi-Object Tracking in Satellite Videos with Graph-Based Multi-Task Modeling

TGraM Multi-Object Tracking in Satellite Videos with Graph-Based Multi-Task Modeling, Qibin He, Xian Sun, Zhiyuan Yan, Beibei Li, Kun Fu Abstract Rece

Qibin He 6 Nov 25, 2022
[CVPR'21] Multi-Modal Fusion Transformer for End-to-End Autonomous Driving

TransFuser This repository contains the code for the CVPR 2021 paper Multi-Modal Fusion Transformer for End-to-End Autonomous Driving. If you find our

695 Jan 05, 2023
ESL: Event-based Structured Light

ESL: Event-based Structured Light Video (click on the image) This is the code for the 2021 3DV paper ESL: Event-based Structured Light by Manasi Mugli

Robotics and Perception Group 29 Oct 24, 2022