Deep Poselets for Human Detection

TL;DR

This paper introduces a novel human detection method using poselets and CNNs, achieving state-of-the-art results by leveraging weakly labeled data and a new feature representation called PDF vectors.

Contribution

It presents a bootstrapping approach to gather large-scale poselet data and a CNN-based pose discriminative feature for improved human detection.

Findings

Achieved state-of-the-art performance on PASCAL datasets.

Developed a compact 256-dimensional PDF feature vector.

Combined poselet and object-level CNNs for enhanced detection.

Abstract

We address the problem of detecting people in natural scenes using a part approach based on poselets. We propose a bootstrapping method that allows us to collect millions of weakly labeled examples for each poselet type. We use these examples to train a Convolutional Neural Net to discriminate different poselet types and separate them from the background class. We then use the trained CNN as a way to represent poselet patches with a Pose Discriminative Feature (PDF) vector -- a compact 256-dimensional feature vector that is effective at discriminating pose from appearance. We train the poselet model on top of PDF features and combine them with object-level CNNs for detection and bounding box prediction. The resulting model leads to state-of-the-art performance for human detection on the PASCAL datasets.