Logo Lanfrica
  • Home
  • Atlas
  • Insights
  • Docs
  • Sign in

© 2026 Lanfrica. All rights reserved. All copyrights of the resources shown on the Lanfrica website belong to the original copyright holders, unless explicitly stated otherwise.

VaradBhagwat5/structure-aware-medical-diffusion

Domain:

healthcare

Record type:

model
Creator:
Var
Host:
A structure-aware conditional diffusion framework for enhancing degraded medical images in low-resource radiology settings, with explicit structural preservation constraints and comparison against GAN-based methods. # Teammate C — Backend + Frontend ## Project structure ``` api/ main.py ← FastAPI app, /enhance endpoint infer.py ← ESRGAN / Pix2Pix inference wrapper requirements.txt frontend/src/app/ services/enhancement.service.ts ← HTTP client components/enhancer/ enhancer.component.ts ← logic enhancer.component.html ← UI: upload · slider · heatmap enhancer.component.scss ← dark-theme styles app.module.ts ``` --- ## Backend setup ```bash cd api python -m venv .venv && source .venv/bin/activate pip install -r requirements.txt uvicorn main:app --reload --port 8000 ``` Health check → localhost ### Connecting real model checkpoints Open `infer.py` and replace the `TODO` lines with your actual checkpoint paths: ```python # ESRGAN m.load_state_dict(torch.load("checkpoints/esrgan_x4.pth", map_location=self.device)) # Pix2Pix m.load_state_dict(torch.load("checkpoints/pix2pix.pth", map_location=self.device)) ``` --- ## Frontend setup ```bash cd frontend npm install ng serve # dev server → localhost ``` ### UI features | Feature | Description | |---|---| | Drag-and-drop upload | Drop zone or click-to-browse | | Model selector | ESRGAN (super-res) · Pix2Pix (translation) | | Scale buttons | ×2 / ×4 / ×8 (ESRGAN) | | Strength slider | Blend 0–100% enhancement | | Before/after comparison | Draggable divider slider | | Attention heatmap | Toggle canvas overlay | | Download | Save enhanced PNG | --- ## API reference ### `POST /enhance` | Field | Type | Default | Description | |---|---|---|---| | `file` | image/* | — | Input image (max 20 MB) | | `model` | string | `esrgan` | `esrgan` or `pix2pix` | | `scale` | int | `4` | Upscale factor (ESRGAN only) | | `strength` | float | `1.0` | Enhancement blend (0.0–1.0) | **Response** ```json { "enhanced_image": " ", "heatmap": " ", "model": "esrgan", "scale": 4, "inference_time_s": 0.342, "original_size": [512, 512], "enhanced …

Visit

github.com

Tasks

computer vision

Licenses

MIT

Similar

Segmentation-Aware Latent Diffusion for Satellite Image Super-Resolution: Enabling Smallholder Farm Boundary DelineationSocial Network Structure and Climate Change Adaptation Information Diffusion among Dairy Farmers: Evidence from Northern TunisiaBengaliMCQ: Structure-Aware Retrieval-Augmented Generation for MCQ Generation and Answer Prediction in a Low-Resource LanguagePopulation-Structure-Aware Machine Learning for Antimicrobial Resistance Prediction from Bacterial Accessory Genomes: A Benchmark on Escherichia coli Ciprofloxacin ResistanceBoluAW/weather-awareContext-aware software

Segmentation-Aware Latent Diffusion for Satellite Image Super-Resolution: Enabling Smallholder Farm Boundary Delineation

Delineating farm boundaries through segmentation of satellite images is a fundamental step in many a

Social Network Structure and Climate Change Adaptation Information Diffusion among Dairy Farmers: Evidence from Northern Tunisia

Climate adaptation information network of 239 dairy farmers in Bizerte Governorate, northern Tunisia

BengaliMCQ: Structure-Aware Retrieval-Augmented Generation for MCQ Generation and Answer Prediction in a Low-Resource Language

Traditional retrieval-augmented generation (RAG) frameworks process documents without attending to t

Population-Structure-Aware Machine Learning for Antimicrobial Resistance Prediction from Bacterial Accessory Genomes: A Benchmark on Escherichia coli Ciprofloxacin Resistance

Abstract

Antimicrobial res

BoluAW/weather-aware

weather-integrated event scheduling platform designed to solve the problem of weather-related event

Context-aware software

With the advent of PDAs (Personal Digital Assistants), smart phones, and other forms of mobile and u