Looking Beyond Appearances: Synthetic Training Data for Deep CNNs in Re-identification

Abstract

Re-identification is generally carried out by encoding the appearance of a subject in terms of outfit, suggesting scenarios where people do not change their attire. In this paper we overcome this restriction, by proposing a framework based on a deep convolutional neural network, SOMAnet, that additionally models other discriminative aspects, namely, structural attributes of the human figure (e.g. height, obesity, gender). Our method is unique in many respects. First, SOMAnet is based on the Inception architecture, departing from the usual siamese framework. This spares expensive data preparation (pairing images across cameras) and allows the understanding of what the network learned. Second, and most notably, the training data consists of a synthetic 100K instance dataset, SOMAset, created by photorealistic human body generation software. Synthetic data represents a good compromise between realistic imagery, usually not required in re-identification since surveillance cameras capture low-resolution silhouettes, and complete control of the samples, which is useful in order to customize the data w.r.t. the surveillance scenario at-hand, e.g. ethnicity. SOMAnet, trained on SOMAset and fine-tuned on recent re-identification benchmarks, outperforms all competitors, matching subjects even with different apparel. The combination of synthetic data with Inception architectures opens up new research avenues in re-identification.

14 pages

References in corpus (9)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Deep Learning Face Representation by Joint Identification-Verification
• Understanding Neural Networks Through Deep Visualization
• Gated Siamese Convolutional Neural Network Architecture for Human Re-Identification
• A Siamese Long Short-Term Memory Architecture for Human Re-Identification
• Learning to rank in person re-identification with metric ensembles
• Learning Classifiers from Synthetic Data Using a Multichannel Autoencoder

Cited by in corpus (42)

• In Defense of the Triplet Loss for Person Re-Identification
• Improving Person Re-identification by Attribute and Identity Learning
• Pedestrian Alignment Network for Large-scale Person Re-identification
• AlignedReID: Surpassing Human-Level Performance in Person Re-Identification
• Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro
• Beyond Part Models: Person Retrieval with Refined Part Pooling (and a Strong Convolutional Baseline)
• Margin Sample Mining Loss: A Deep Learning Based Method for Person Re-identification
• SVDNet for Pedestrian Retrieval
• A survey of synthetic data augmentation methods in computer vision
• Second-order Non-local Attention Networks for Person Re-identification
• Three Dimensional Fluorescence Microscopy Image Synthesis and Segmentation
• Parameter-Efficient Person Re-identification in the 3D Space
• Let Features Decide for Themselves: Feature Mask Network for Person Re-identification
• Learning Multi-Human Optical Flow
• Pyramidal Person Re-IDentification via Multi-Loss Dynamic Training
• Batch DropBlock Network for Person Re-identification and Beyond
• ClearGrasp: 3D Shape Estimation of Transparent Objects for Manipulation
• Locality Aware Appearance Metric for Multi-Target Multi-Camera Tracking
• Person Recognition in Personal Photo Collections
• Dissecting Person Re-identification from the Viewpoint of Viewpoint
• When Person Re-identification Meets Changing Clothes
• Human Image Generation: A Comprehensive Survey
• An Evaluation of Deep CNN Baselines for Scene-Independent Person Re-Identification
• Surpassing Real-World Source Training Data: Random 3D Characters for Generalizable Person Re-Identification
• Deep Feature Learning via Structured Graph Laplacian Embedding for Person Re-Identification
• Discovering Underlying Person Structure Pattern with Relative Local Distance for Person Re-identification
• UnrealPerson: An Adaptive Pipeline towards Costless Person Re-identification
• Exploring Uncertainty in Conditional Multi-Modal Retrieval Systems
• Effective Image Retrieval via Multilinear Multi-index Fusion
• Learning to simulate complex scenes
• Attribute analysis with synthetic dataset for person re-identification
• Less is More: Learning from Synthetic Data with Fine-grained Attributes for Person Re-Identification
• Unsupervised Domain Adaptive Learning via Synthetic Data for Person Re-identification
• Texel-Att: Representing and Classifying Element-based Textures by Attributes
• A Masked Face Classification Benchmark on Low-Resolution Surveillance Images
• SyDog: A Synthetic Dog Dataset for Improved 2D Pose Estimation
• Sparse Label Smoothing Regularization for Person Re-Identification
• Deep Person Re-Identification with Improved Embedding and Efficient Training
• Learning Deep Representations by Mutual Information for Person Re-identification
• From Real to Synthetic and Back: Synthesizing Training Data for Multi-Person Scene Understanding
• Imbalance Robust Softmax for Deep Embeeding Learning
• Taking A Closer Look at Synthesis: Fine-grained Attribute Analysis for Person Re-Identification