Expecting the Unexpected: Training Detectors for Unusual Pedestrians with Adversarial Imposters

By Shiyu Huang, Deva Ramanan. 2017

54 citations recorded by the source corpus.

As autonomous vehicles become an every-day reality, high-accuracy pedestrian detection is of paramount practical importance. Pedestrian detection is a highly researched topic with mature methods, but most datasets (for both training and evaluation) focus on common scenes of people engaged in typical walking poses on sidewalks. But performance is most crucial for dangerous scenarios that are rarely observed, such as children playing in the street and people using bicycles/skateboards in unexpected ways. Such in-the-tail data is notoriously hard to observe, making both training and testing difficult. To analyze this problem, we have collected a novel annotated dataset of dangerous scenarios called the Precarious Pedestrian dataset. Even given a dedicated collection effort, it is relatively small by contemporary standards (≈ 1000 images). To explore large-scale data-driven learning, we explore the use of synthetic data generated by a game engine. A significant challenge is selected the right priors or parameters for synthesis: we would like realistic data with realistic poses and object configurations. Inspired by Generative Adversarial Networks, we generate a massive amount of synthetic data and train a discriminative classifier to select a realistic subset (that fools the classifier), which we deem Synthetic Imposters. We demonstrate that this pipeline allows one to generate realistic (or adverserial) training data by making use of rendering/animation engines. Interestingly, we also demonstrate that such data can be used to rank algorithms, suggesting that Synthetic Imposters can also be used for in-the-tail validation at test-time, a notoriously difficult challenge for real-world deployment.

Subfields: Automotive Engineering, Computer Vision and Pattern Recognition, Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Automotive Engineering

In Computer Vision and Pattern Recognition

In Safety, Robustness & Interpretability

Citations in the corpus

Bidirectional Conditional Generative Adversarial Networks (2017, methodology): “data from complex distributions for a variety of applications such as image editing [23], image synthesis from text descriptions [24], image super-resolution [14], video summarization [17] and others [2,8,10,11,15,26,28,29,30, 32,34,35], with very promising results. The core methodology for modeling”
EuroCity Persons: A Novel Benchmark for Person Detection in Traffic Scenes (2018, methodology): “For Faster R-CNN, R-FCN and SSD we use VGG-16 [45] as base network as it is very common in pedestrian detection [4], [60], [62], [63], [71].”