AIbase
EN
Home
Resnet 50 FV2 Finetuned Memes
R

Resnet 50 FV2 Finetuned Memes

Developed by jayanta
A meme classification model fine-tuned based on Microsoft's ResNet-50 architecture, achieving 64.5% accuracy on image classification tasks
Image Classification
Transformers
Open Source License:Apache-2.0#Meme Classification#ResNet Fine-tuning#Multi-metric Optimization
Downloads 24
Release Time : 10/21/2022

Model Overview

This is a convolutional neural network optimized for meme image classification tasks, fine-tuned through transfer learning on the ResNet-50 foundation

Model Features

Transfer Learning Optimization
Fine-tuned based on the mature ResNet-50 architecture, effectively utilizing the visual feature extraction capabilities of the pre-trained model
Multi-metric Evaluation
Provides comprehensive evaluation metrics including accuracy, precision, recall and F1 score
Standardized Training Process
Employs best practices such as linear learning rate scheduling and warmup for model optimization

Model Capabilities

Image Classification
Meme Recognition
Visual Feature Extraction

Use Cases

Social Media
Automatic Meme Classification
Automatically classify and manage memes uploaded by users
Achieved 64.5% classification accuracy on test set
Content Moderation
Inappropriate Content Identification
Identify memes that may contain inappropriate content
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