Pre-training for Action Recognition with Automatically Generated Fractal Datasets

TL;DR

This paper introduces a novel method for pre-training action recognition models using large-scale synthetic video datasets generated through fractal geometry, reducing reliance on real data and improving downstream performance.

Contribution

It presents a new approach to generate diverse synthetic videos with fractal geometry for pre-training, emulating real video properties to enhance action recognition tasks.

Findings

Synthetic pre-training approaches achieve results close to or better than Kinetics-based pre-training.

Generated videos exhibit high variety and complex multi-scale structures.

Pre-training with synthetic videos improves performance on multiple action recognition benchmarks.

Abstract

In recent years, interest in synthetic data has grown, particularly in the context of pre-training the image modality to support a range of computer vision tasks, including object classification, medical imaging etc. Previous work has demonstrated that synthetic samples, automatically produced by various generative processes, can replace real counterparts and yield strong visual representations. This approach resolves issues associated with real data such as collection and labeling costs, copyright and privacy. We extend this trend to the video domain applying it to the task of action recognition. Employing fractal geometry, we present methods to automatically produce large-scale datasets of short synthetic video clips, which can be utilized for pre-training neural models. The generated video clips are characterized by notable variety, stemmed by the innate ability of fractals to generate complex multi-scale structures. To narrow the domain gap, we further identify key properties of real videos and carefully emulate them during pre-training. Through thorough ablations, we determine the attributes that strengthen downstream results and offer general guidelines for pre-training with synthetic videos. The proposed approach is evaluated by fine-tuning pre-trained models on established action recognition datasets HMDB51 and UCF101 as well as four other video benchmarks related to group action recognition, fine-grained action recognition and dynamic scenes. Compared to standard Kinetics pre-training, our reported results come close and are even superior on a portion of downstream datasets. Code and samples of synthetic videos are available at https://github.com/davidsvy/fractal_video .