Simulating Content Consistent Vehicle Datasets with Attribute Descent

Abstract

This paper uses a graphic engine to simulate a large amount of training data with free annotations. Between synthetic and real data, there is a two-level domain gap, i.e., content level and appearance level. While the latter has been widely studied, we focus on reducing the content gap in attributes like illumination and viewpoint. To reduce the problem complexity, we choose a smaller and more controllable application, vehicle re-identification (re-ID). We introduce a large-scale synthetic dataset VehicleX. Created in Unity, it contains 1,362 vehicles of various 3D models with fully editable attributes. We propose an attribute descent approach to let VehicleX approximate the attributes in real-world datasets. Specifically, we manipulate each attribute in VehicleX, aiming to minimize the discrepancy between VehicleX and real data in terms of the Fréchet Inception Distance (FID). This attribute descent algorithm allows content domain adaptation (DA) orthogonal to existing appearance DA methods. We mix the optimized VehicleX data with real-world vehicle re-ID datasets, and observe consistent improvement. With the augmented datasets, we report competitive accuracy. We make the dataset, engine and our codes available at https://github.com/yorkeyao/VehicleX.

arXiv commentAccepted to ECCV2020

References in corpus (8)

• GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
• In Defense of the Triplet Loss for Person Re-Identification
• Semantic Foggy Scene Understanding with Synthetic Data
• Unity: A General Platform for Intelligent Agents
• AI2-THOR: An Interactive 3D Environment for Visual AI
• Vehicle Re-Identification: an Efficient Baseline Using Triplet Embedding
• Learning data augmentation policies using augmented random search
• Learning to simulate complex scenes

Cited by in corpus (13)

• Self-paced Contrastive Learning with Hybrid Memory for Domain Adaptive Object Re-ID
• VehicleNet: Learning Robust Visual Representation for Vehicle Re-identification
• Parameter-Efficient Person Re-identification in the 3D Space
• VOC-ReID: Vehicle Re-identification based on Vehicle-Orientation-Camera
• Joint Disentangling and Adaptation for Cross-Domain Person Re-Identification
• Multiview Detection with Feature Perspective Transformation
• Multi-Domain Learning and Identity Mining for Vehicle Re-Identification
• Learning to simulate complex scenes
• An Empirical Study of Vehicle Re-Identification on the AI City Challenge
• Pairwise-GAN: Pose-based View Synthesis through Pair-Wise Training
• StRDAN: Synthetic-to-Real Domain Adaptation Network for Vehicle Re-Identification
• Synthetic Data Are as Good as the Real for Association Knowledge Learning in Multi-object Tracking
• Viewpoint-Aware Channel-Wise Attentive Network for Vehicle Re-Identification