V

Videomae Base Finetuned

Developed by sheraz179
A video understanding model fine-tuned on an unknown dataset based on MCG-NJU/videomae-base, achieving an F1 score of 0.7147
Downloads 15
Release Time : 2/8/2023

Model Overview

This model is a fine-tuned version of the VideoMAE base architecture, focusing on video content understanding tasks, potentially applicable to scenarios such as video classification or action recognition

Model Features

Efficient Video Representation Learning
Based on the VideoMAE architecture, effectively learns spatiotemporal features of videos through masked autoencoder pre-training
Excellent Fine-tuning Performance
Achieves an F1 score of 0.7147 on the evaluation set, indicating strong discriminative capability
Lightweight Training
Efficient fine-tuning with a small batch size (3)

Model Capabilities

Video Feature Extraction
Video Content Understanding
Spatiotemporal Pattern Recognition

Use Cases

Video Analysis
Action Recognition
Identify human actions or activities in videos
F1 score 0.7147
Video Classification
Classify and label video content
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase