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

Published in ECCV, 2020

Recommended Citation: Khorramshahi P, Peri N, Chen JC, Chellappa R. The Devil is in the Details: Self-Sueprvised Attention for Vehicle Re-ID. In: European Conference on Computer Vision, ECCV (2020)

In recent years, the research community has approached the problem of vehicle re-identi cation (re-id) with attention-based models, speci cally focusing on regions of a vehicle containing discriminative information. These re-id methods rely on expensive key-point labels, part annotations, and additional attributes including vehicle make, model, and color. Given the large number of vehicle re-id datasets with various levels of annotations, strongly-supervised methods are unable to scale across different domains. In this paper, we present Self-supervised Attention for Vehicle Re-identi cation (SAVER), a novel approach to effectively learn vehicle-speci c discriminative features. Through extensive experimentation, we show that SAVER improves upon the state-of-the-art on challenging vehicle re-id benchmarks including Veri-776, VehicleID, Vehicle-1M and Veri-Wild. SAVER demonstrates how proper regularization techniques signi cantly constrain the vehicle re-id task and help generate robust deep features.

[Download Paper]