GID-segmentation-FPN_resnet34 Open-source Image Segmentation Model

GID Segmentation FPN Resnet34

Developed by jiabing24

FPN-based image segmentation model with ResNet34 encoder, optimized for GID dataset

Open Source License:MIT #Multispectral image segmentation #High-precision semantic segmentation #Remote sensing image processing

Downloads 16

Release Time : 7/9/2024

Model Overview

This is a semantic segmentation model using the FPN architecture, primarily designed for land cover classification tasks in remote sensing images. The model employs ResNet34 as the encoder, supporting 4-channel input and single-category output.

Model Features

Multi-scale feature fusion

Utilizes the FPN architecture to achieve top-down and lateral connections in the feature pyramid, effectively integrating features at different scales

Pre-trained encoder

Uses ResNet34 pre-trained on ImageNet as the encoder to enhance feature extraction capabilities

4-channel input support

Supports 4-channel remote sensing image input, tailored for the GID dataset

Model Capabilities

Remote sensing image segmentation

Land cover classification

Multispectral image analysis

Use Cases

Remote sensing analysis

Land cover classification

Classifies and segments different land types in remote sensing images

Achieved IoU of 0.761 on the test set

Property	Details
Dataset Name	GID

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

GID Segmentation FPN Resnet34

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 FPN Model Card

🚀 Quick Start

Load trained model

✨ Features

Model init parameters

Model metrics

Dataset

📚 Documentation

📄 License