AIbase
EN
Home
Oxford Pet Segmentation
O

Oxford Pet Segmentation

Developed by castiello
PyTorch-based FPN image segmentation model supporting multiple encoder architectures, suitable for semantic segmentation tasks
Image Segmentation
Safetensors
Open Source License:MIT#Pet Image Segmentation#ResNet Encoder#High IoU Accuracy
Downloads 34
Release Time : 4/17/2025

Model Overview

This model implements the Feature Pyramid Network (FPN) architecture for efficient image semantic segmentation. Supports various pre-trained encoders and flexible decoder parameter configurations, suitable for segmentation tasks in different scenarios.

Model Features

Flexible Encoder Selection
Supports multiple pre-trained encoders (e.g., ResNet34), allowing easy switching between different backbone networks
Configurable Decoder
Provides rich decoder parameter configuration options, including channel count, dropout rate, and interpolation methods
Pre-trained Weight Support
Encoders can utilize ImageNet pre-trained weights to accelerate model convergence

Model Capabilities

Image Semantic Segmentation
Feature Pyramid Processing
Multi-scale Feature Fusion

Use Cases

Computer Vision
Pet Image Segmentation
Accurate segmentation of pets and background on the Oxford Pets dataset
Test IoU reached 0.914
Medical Image Analysis
Can be used for organ or lesion area segmentation in medical images
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase