基于AI的图片/视频硬字幕去除、文本水印去除，无损分辨率生成去字幕、去水印后的图片/视频文件。无需申请第三方API，本地实现。AI-based tool for removing hard-coded subtitles and text-like watermarks from videos or Pictures.

A collection of state-of-the-art video or single-image super-resolution architectures, reimplemented in tensorflow.

Efficient & Generic Video Super-Resolution

Latest development of ISR/VSR. Papers and related resources, mainly state-of-the-art and novel works in ICCV, ECCV and CVPR about image super-resolution and video super-resolution.

It is a re-implementation of paper named "Deep Video Super-Resolution Network Using Dynamic Upsampling Filters Without Explicit Motion Compensation" called VSR-DUF model. There are both training codes and test codes about VSR-DUF based tensorflow.

The state-of-the-art VSR

Official repository containing code and other material from the paper "Efficient Video Super-Resolution through Recurrent Latent Space Propagation" (https://arxiv.org/abs/1909.08080).

SyncVSR: Data-Efficient Visual Speech Recognition with End-to-End Crossmodal Audio Token Synchronization (Interspeech 2024)

This is an official implementation of Video Super-Resolution via a Spatio-Temporal Alignment Network.

Nokia PCE Segment Routing testing

Grounding Language Models for Compositional and Spatial Reasoning

ICIP2024 challenge for 360 super resolution

ODVista dataset for omnidirectional video super-resolution and quality enhancement tasks.

A configuration-driven model training framework built specifically to train high quality image & video generator networks in Pytorch.

HiRN: Hierarchical Recurrent Neural Network for Video Super-Resolution (VSR) using Two-Stage Feature Evolution - Official Repository (Applied Soft Computing)

Speaker-Independent Speech Recognition using Visual Features

Group-based Bi-Directional Recurrent Wavelet Neural Network for Efficient Video Super-Resolution (VSR) - Official Repository (Pattern Recognition Letters)

Firmware for the VSR as known from the Delorean Time Machine

HiRN: Hierarchical Recurrent Neural Network for Video Super-Resolution (VSR) using Two-Stage Feature Evolution - Official Repository (Applied Soft Computing)

Group-based Bi-Directional Recurrent Wavelet Neural Network for Efficient Video Super-Resolution (VSR) - Official Repository (Pattern Recognition Letters)