Exploiting Image-trained CNN Architectures for Unconstrained Video Classification

TL;DR

This paper explores how to effectively use image-trained CNNs for unconstrained video classification, demonstrating significant performance improvements through optimized strategies and feature fusion.

Contribution

It systematically studies pooling, normalization, layer selection, and fusion techniques to enhance CNN-based video event detection, achieving state-of-the-art results.

Findings

Optimized CNN feature extraction improves performance.

Late fusion of CNN and motion features boosts accuracy.

Achieves state-of-the-art results on UCF-101 dataset.

Abstract

We conduct an in-depth exploration of different strategies for doing event detection in videos using convolutional neural networks (CNNs) trained for image classification. We study different ways of performing spatial and temporal pooling, feature normalization, choice of CNN layers as well as choice of classifiers. Making judicious choices along these dimensions led to a very significant increase in performance over more naive approaches that have been used till now. We evaluate our approach on the challenging TRECVID MED'14 dataset with two popular CNN architectures pretrained on ImageNet. On this MED'14 dataset, our methods, based entirely on image-trained CNN features, can outperform several state-of-the-art non-CNN models. Our proposed late fusion of CNN- and motion-based features can further increase the mean average precision (mAP) on MED'14 from 34.95% to 38.74%. The fusion approach achieves the state-of-the-art classification performance on the challenging UCF-101 dataset.