How important are Deformable Parts in the Deformable Parts Model?

TL;DR

This paper investigates the relative importance of deformable parts in the Deformable Parts Model, revealing that increasing components and appearance-based clustering can outperform deformable parts, which can be turned off without significant performance loss.

Contribution

The study demonstrates that multiple components and appearance-based clustering are more crucial than deformable parts in the DPM, challenging previous assumptions.

Findings

Increasing components improves detection performance.

Part deformations can be eliminated with minimal performance loss.

Multiple components benefit scene classification tasks.

Abstract

The main stated contribution of the Deformable Parts Model (DPM) detector of Felzenszwalb et al. (over the Histogram-of-Oriented-Gradients approach of Dalal and Triggs) is the use of deformable parts. A secondary contribution is the latent discriminative learning. Tertiary is the use of multiple components. A common belief in the vision community (including ours, before this study) is that their ordering of contributions reflects the performance of detector in practice. However, what we have experimentally found is that the ordering of importance might actually be the reverse. First, we show that by increasing the number of components, and switching the initialization step from their aspect-ratio, left-right flipping heuristics to appearance-based clustering, considerable improvement in performance is obtained. But more intriguingly, we show that with these new components, the part deformations can now be completely switched off, yet obtaining results that are almost on par with the original DPM detector. Finally, we also show initial results for using multiple components on a different problem -- scene classification, suggesting that this idea might have wider applications in addition to object detection.