TecoGAN ist ein Generative Adversarial Network (GAN), das für Video-Super-Resolution entwickelt wurde. Es fungiert als spatio-temporaler Video-Upscaler, der die Auflösung von Videosequenzen erhöht und gleichzeitig qualitativ hochwertige Bilder aus niedrig aufgelösten Eingaben rekonstruiert.
Die Hauptfunktionen von thunil/tecogan sind: Video Super-Resolution Suites, Generative Adversarial Networks, Video Resolution Reconstruction, GAN-Based Video Super-Resolution Models, Temporal Coherence Propagations, Temporal Coherence Frameworks, Temporal Consistency Optimization, Spatio-Temporal Upscalers.
Open-Source-Alternativen zu thunil/tecogan sind unter anderem: numz/comfyui-seedvr2_videoupscaler — ComfyUI-SeedVR2_VideoUpscaler is an AI video upscaling tool that uses diffusion models to increase the resolution of… aladdinpersson/machine-learning-collection — This project is a machine learning educational repository providing a collection of implementations and guides for… goodfeli/adversarial — This project is a generative adversarial network implementation and research framework. It provides the tools and… wiseodd/generative-models — This is a generative AI model library containing a collection of PyTorch and TensorFlow implementations for creating… paddlepaddle/paddlegan — PaddleGAN is a generative AI framework and deep learning computer vision library built on the PaddlePaddle framework.… david-gpu/srez — Srez is a deep learning image super-resolution framework designed to upscale low-resolution images into sharp,…
ComfyUI-SeedVR2_VideoUpscaler is an AI video upscaling tool that uses diffusion models to increase the resolution of videos and images while maintaining visual consistency across frames. The project implements distributed video rendering by splitting datasets into chunks for parallel processing across multiple GPUs. It utilizes model compilation and specialized attention backends to reduce inference latency and increase throughput. Additional capabilities include video color correction using wavelet and LAB matching methods to preserve color fidelity. Hardware memory is managed through block
This project is a machine learning educational repository providing a collection of implementations and guides for machine learning and deep learning algorithms. It serves as a deep learning model library and a reference for training workflows, covering foundational machine learning, convolutional, recurrent, and transformer architectures. The collection includes a generative adversarial network suite for synthesizing realistic images and performing image-to-image translation. It also functions as a computer vision implementation guide for object detection and semantic segmentation, alongside
This project is a generative adversarial network implementation and research framework. It provides the tools and hyperparameters necessary to train and evaluate generative models across various datasets, specifically designed to reproduce results from academic research. The framework includes a Parzen density likelihood estimator to calculate model log likelihood. This allows for the quantitative evaluation of generative distributions and the measurement of overall model performance. The codebase covers machine learning research capabilities, focusing on the training of adversarial networks
This is a generative AI model library containing a collection of PyTorch and TensorFlow implementations for creating synthetic data and modeling complex probability distributions. It serves as a multi-framework repository of deep learning models designed for learning and replicating data patterns. The project provides specialized implementation suites for several generative architectures. This includes Generative Adversarial Networks using competing generator and discriminator models, Variational Autoencoder frameworks that map data to a latent space, and Restricted Boltzmann Machine and Deep