# CVNet: Lightweight Cross-View Vehicle ReID with Multi-Scale Localization

## TL;DR

CVNet is a lightweight network for cross-view vehicle re-identification that uses efficient modules to handle scale variations and improve performance on edge devices.

## Contribution

The novel contribution is CVNet with MSL and DFC modules, and the release of the CVPair v1.0 dataset for cross-view ReID.

## Key findings

- CVNet achieves state-of-the-art performance on CVPair v1.0, VehicleID, and VeRi776 benchmarks.
- The MSL module effectively localizes salient regions and handles viewpoint variations.
- The DFC module captures both unique and shared cross-view features efficiently.

## Abstract

Cross-view vehicle re-identification (ReID) between aerial and ground perspectives is challenging due to limited computational resources on edge devices and significant scale variations. We propose CVNet, a lightweight network with two key modules: the multi-scale localization (MSL) module and the deep–shallow filtrate collaboration (DFC) module. The MSL module employs multi-scale depthwise separable convolutions and a localization attention mechanism to extract multi-scale features and localize salient regions, addressing viewpoint variations. DFC employs a dual-branch design comprising deep and shallow branches, integrating a filtration module optimized via neural architecture search, a collaboration module, and lightweight convolutions. This design effectively captures both unique and shared cross-view features, ensuring efficient and robust feature representation. We also release a new CVPair v1.0 dataset, the first benchmark for cross-view ReID, containing 14,969 images of 894 vehicle identities, offering results of traditional and lightweight methods. CVNet achieves state-of-the-art performance on CVPair v1.0, VehicleID, and VeRi776, advancing cross-view vehicle ReID. The dataset will be released publicly.