PAMTRI: Pose-Aware Multi-Task Learning for Vehicle Re-Identification Using Highly Randomized Synthetic Data

  title={PAMTRI: Pose-Aware Multi-Task Learning for Vehicle Re-Identification Using Highly Randomized Synthetic Data},
  author={Zheng Tang and Milind R. Naphade and Stan Birchfield and Jonathan Tremblay and William Hodge and Ratnesh Kumar and Shuo Wang and Xiaodong Yang},
  journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
In comparison with person re-identification (ReID), which has been widely studied in the research community, vehicle ReID has received less attention. Vehicle ReID is challenging due to 1) high intra-class variability (caused by the dependency of shape and appearance on viewpoint), and 2) small inter-class variability (caused by the similarity in shape and appearance between vehicles produced by different manufacturers). To address these challenges, we propose a Pose-Aware Multi-Task Re… 

Figures and Tables from this paper

Attributes Guided Feature Learning for Vehicle Re-Identification

This paper proposes a novel deep network architecture, which guided by meaningful attributes including camera views, vehicle types and colors for vehicle Re-ID, and is end-to-end trained and contains three subnetworks of deep features embedded by the corresponding attributes.

VOC-RelD: Vehicle Re-identification based on Vehicle-Orientation-Camera

This work focuses on the failure cases caused by similar background and shape, and proposes an approach named VOC- ReID, taking the triplet vehicle-orientation-camera as a whole and reforming background/shape similarity as camera/orientation re-identification.

Large Scale Vehicle Re-Identification by Knowledge Transfer from Simulated Data and Temporal Attention

This work addresses the re-id task by relying on well-proven design choices from the closely related person re-identification literature, and explicitly address viewpoint and occlusions variation.

AttributeNet: Attribute Enhanced Vehicle Re-Identification

Orientation-aware Vehicle Re-identification with Semantics-guided Part Attention Network

This work proposes a dedicated Semantics-guided Part Attention Network (SPAN) to robustly predict part attention masks for different views of vehicles given only image-level semantic labels during training and introduces Co-occurrence Part-attentive Distance Metric (CPDM) which places greater emphasis on co-occurring vehicle parts when evaluating the feature distance of two images.

Model Latent Views With Multi-Center Metric Learning for Vehicle Re-Identification

A multi-center metric learning framework for multi-view vehicle Re-ID that model latent views from vehicle visual appearance directly without any extra labels except ID is proposed and shows the superiority of the proposed framework in contrast to a series of existing state-of-the-arts benchmarks.

The Devil is in the Details: Self-Supervised Attention for Vehicle Re-Identification

Self-supervised Attention for Vehicle Re-identification (SAVER) is presented, a novel approach to effectively learn vehicle-specific discriminative features and improves upon the state-of-the-art on challenging VeRi, VehicleID, Vehicle-1M and VERI-Wild datasets.

Combining Pose Invariant and Discriminative Features for Vehicle Reidentification

PIF and PDF are complementary and that PRF produces competitive performance compared with state-of-the-art approaches and can alleviate the influence of the pose barrier problem.

Beyond the Parts: Learning Multi-view Cross-part Correlation for Vehicle Re-identification

The PCRNet explores vehicle parsing to learn discriminative part-level features, model the correlation among vehicle parts, and achieve precise part alignment for vehicle Re-Id, and proposes a self-supervised part prediction loss to make the GCNs generate features of invisible parts from visible parts under different viewpoints.

Region-aware multi-resolution learning for vehicle re-identification using mask

A region-aware multi-resolution (RAMR) Re-ID framework that can extract features from a series of local regions without extra manual annotations is proposed and indicates that the RAMR model achieves significant improvement in comparison with other state-of-the-art methods.



Group-Sensitive Triplet Embedding for Vehicle Reidentification

A deep metric learning method, group-sensitive-triplet embedding (GS-TRE), to recognize and retrieve vehicles, in which intraclass variance is elegantly modeled by incorporating an intermediate representation “group” between samples and each individual vehicle in the triplet network learning.

Vehicle Re-identification: an Efficient Baseline Using Triplet Embedding

The approach is simpler and more straightforward for training utilizing only identity-level annotations, along with one of the smallest published embedding dimensions for efficient inference, and a formal evaluation of a triplet sampling variant (batch sample) into the re-identification literature.

Vehicle Re-Identification by Adversarial Bi-Directional LSTM Network

  • Yi ZhouL. Shao
  • Computer Science
    2018 IEEE Winter Conference on Applications of Computer Vision (WACV)
  • 2018
This paper proposes an Adversarial Bi-directional LSTM Network (ABLN), a global vehicle representation containing all views' information can be inferred from only one visible view, and then used for learning to measure the distance between two vehicles with arbitrary views.

PROVID: Progressive and Multimodal Vehicle Reidentification for Large-Scale Urban Surveillance

This paper proposes PROVID, a PROgressive Vehicle re-IDentification framework based on deep neural networks, which not only utilizes the multimodality data in large-scale video surveillance, such as visual features, license plates, camera locations, and contextual information, but also considers vehicle reidentification in two progressive procedures: coarse- to-fine search in the feature domain, and near-to-distantsearch in the physical space.

Learning Deep Neural Networks for Vehicle Re-ID with Visual-spatio-Temporal Path Proposals

A two-stage framework that incorporates complex spatio-temporal information for effectively regularizing the re-identification results is proposed and extensive experiments and analysis show the effectiveness of the proposed method and individual components.

Deep Relative Distance Learning: Tell the Difference between Similar Vehicles

A Deep Relative Distance Learning (DRDL) method is proposed which exploits a two-branch deep convolutional network to project raw vehicle images into an Euclidean space where distance can be directly used to measure the similarity of arbitrary two vehicles.

Joint Discriminative and Generative Learning for Person Re-Identification

This paper proposes a joint learning framework that couples re-id learning and data generation end-to-end and renders significant improvement over the baseline without using generated data, leading to the state-of-the-art performance on several benchmark datasets.

Single-Camera and Inter-Camera Vehicle Tracking and 3D Speed Estimation Based on Fusion of Visual and Semantic Features

A histogram-based adaptive appearance model is introduced to learn long-term history of visual features for each vehicle target and evolutionary optimization is applied to camera calibration for reliable 3D speed estimation.

Scalable Person Re-identification: A Benchmark

A minor contribution, inspired by recent advances in large-scale image search, an unsupervised Bag-of-Words descriptor is proposed that yields competitive accuracy on VIPeR, CUHK03, and Market-1501 datasets, and is scalable on the large- scale 500k dataset.

Exploiting Multi-grain Ranking Constraints for Precisely Searching Visually-similar Vehicles

This paper model the relationship of vehicle images as multiple grains, and proposes two approaches to alleviate the precise vehicle search problem by exploiting multi-grain ranking constraints, which achieve the state-of-the-art performance on both datasets.