OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

Spsr

Spsr is a Python and PyTorch research implementation of gradient-guided, structure-preserving image super-resolution. It includes scripts for training and testing models, dataset preprocessing code, and an evaluation toolbox for image-quality metrics.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Spsr — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Spsr when you need a configurable research codebase for gradient-guided image enlargement, model training or quantitative evaluation, and can work with its Python, PyTorch and CUDA requirements.

Who it is for

  • Computer-vision researchers studying image super-resolution or restoration.
  • Developers who want to configure and run a PyTorch super-resolution model on low-resolution images.
  • Researchers comparing generated images with image-quality metrics.

Documented uses

  • Generate super-resolution images from a folder of low-resolution inputs.
  • Prepare image datasets and train an SPSR model with configurable paths and hyperparameters.
  • Evaluate super-resolution outputs using metrics including PSNR, SSIM, NIQE and LPIPS.
Before you choose

Requirements & limits

  • The README specifies Python 3, PyTorch 1.0 or newer, an NVIDIA GPU and CUDA.
  • Testing requires configuring the input folder and model path in a JSON options file.
  • Training involves preparing paired image data and editing dataset paths and model settings.
  • The repository includes an evaluation toolbox with perceptual and distortion metrics.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Generate super-resolution images with the SPSR model

Configure the documented test workflow to produce higher-resolution images from low-resolution inputs.

Before you start

  • Python 3, PyTorch 1.0 or newer, an NVIDIA GPU with CUDA, and the listed Python packages.
  • A low-resolution image folder and the SPSR model file, spsr.pth, placed in ./experiments/pretrain_models.

Steps

  1. Set the low-resolution image folder in options/test/test_spsr.json.

  2. Set the pretrained-model directory in the test configuration.

  3. Run the command below.

    Documented command
    python test.py -opt options/test/test_spsr.json

Expected outcome

Generated super-resolution images are written to ./results.

Guide limitations

  • The README requires configuration changes for the user's files and model path.
  • The documented setup specifies an NVIDIA GPU and CUDA.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
18 October 2026
Original website title
Maclory/SPSR · Official homepage title · retained in its source language
How we review sourcesSuggest a correction