How to Sample High Quality 3D Fractals for Action Recognition Pre-Training?

2026-02-12

AI Summary

This work proposes an efficient 3D fractal video generation framework based on Iterated Function Systems (IFS) to address the limitations of traditional methods, which suffer from low computational efficiency and degenerate structures that hinder their applicability to pretraining action recognition models. By incorporating temporal transformations to construct temporally coherent fractal videos and introducing a Targeted Smart Filtering sampling strategy, the proposed approach accelerates generation by nearly two orders of magnitude while enhancing structural diversity. This effectively mitigates performance degradation in downstream tasks caused by excessive constraints. Experimental results demonstrate that the method significantly outperforms existing 3D fractal filtering techniques in action recognition pretraining, achieving a superior balance between computational efficiency and representational capacity.

Technology Category

• Computer Vision: 3D Computer Vision
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Representation Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Synthetic datasets are being recognized in the deep learning realm as a valuable alternative to exhaustively labeled real data. One such synthetic data generation method is Formula Driven Supervised Learning (FDSL), which can provide an infinite number of perfectly labeled data through a formula driven approach, such as fractals or contours. FDSL does not have common drawbacks like manual labor, privacy and other ethical concerns. In this work we generate 3D fractals using 3D Iterated Function Systems (IFS) for pre-training an action recognition model. The fractals are temporally transformed to form a video that is used as a pre-training dataset for downstream task of action recognition. We find that standard methods of generating fractals are slow and produce degenerate 3D fractals. Therefore, we systematically explore alternative ways of generating fractals and finds that overly-restrictive approaches, while generating aesthetically pleasing fractals, are detrimental for downstream task performance. We propose a novel method, Targeted Smart Filtering, to address both the generation speed and fractal diversity issue. The method reports roughly 100 times faster sampling speed and achieves superior downstream performance against other 3D fractal filtering methods.

Problem

• 3D fractals
• action recognition
• pre-training
• synthetic data
• fractal generation

Innovation

• Targeted Smart Filtering
• 3D fractals
• action recognition pre-training
• Formula Driven Supervised Learning
• 3D Iterated Function Systems