AIbase
EN
Home
Convnext Tiny 224 Finetuned On Unlabelled IA With Snorkel Labels
C

Convnext Tiny 224 Finetuned On Unlabelled IA With Snorkel Labels

Developed by ImageIN
This is a computer vision model based on the ConvNeXt-Tiny architecture, fine-tuned on unsupervised data using pseudo-labels generated by Snorkel
Image Classification
Transformers
#Snorkel weak supervision learning#ConvNeXt architecture#High-precision classification
Downloads 14
Release Time : 10/13/2022

Model Overview

This model is a variant of the ConvNeXt-Tiny architecture, specifically fine-tuned on unlabeled image data using pseudo-labels generated by the Snorkel framework. Suitable for general image classification tasks.

Model Features

Unsupervised data fine-tuning
Trained on unsupervised data using pseudo-labels generated by the Snorkel framework, reducing dependence on labeled data
Efficient architecture
Based on ConvNeXt-Tiny architecture, maintaining high accuracy while having a compact model size
Robust performance
Balanced performance across multiple evaluation metrics, achieving an F1 score of 0.8058

Model Capabilities

Image classification
Feature extraction
Transfer learning

Use Cases

General image recognition
Product categorization
Classifying product images on e-commerce platforms
Achieves 86.29% accuracy
Content moderation
Identifying inappropriate content in images
F1 score 0.8058
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase