MixVPR: Feature Mixing for Visual Place Recognition

Visual Place Recognition (VPR) is a crucial part of mobile robotics and autonomous driving as well as other computer vision tasks. It refers to the process of identifying a place depicted in a query image using only computer vision. At large scale, repetitive structures, weather and illumination changes pose a real challenge, as appearances can drastically change over time. Along with tackling these challenges, an efficient VPR technique must also be practical in real-world scenarios where latency matters. To address this, we introduce MixVPR, a new holistic feature aggregation technique that takes feature maps from pre-trained backbones as a set of global features. Then, it incorporates a global relationship between elements in each feature map in a cascade of feature mixing, eliminating the need for local or pyramidal aggregation as done in NetVLAD or TransVPR. We demonstrate the effectiveness of our technique through extensive experiments on multiple large-scale benchmarks. Our method outperforms all existing techniques by a large margin while having less than half the number of parameters compared to CosPlace and NetVLAD. We achieve a new all-time high recall@1 score of 94.6% on Pitts250k-test, 88.0% on MapillarySLS, and more importantly, 58.4% on Nordland. Finally, our method outperforms two-stage retrieval techniques such as Patch-NetVLAD, TransVPR and SuperGLUE all while being orders of magnitude faster.

1. Introduction

Visual place recognition (VPR) is an essential part of many robotics [11], [9], [10], [15], [18], [22] and computer vision tasks [2], [23], [27], [16], [17], [45], [6] such as autonomous driving [12], SLAM [49], image geo-localization [38], [7], virtual reality [31] and 3D reconstruction [29]. A visual place recognition system retrieves the location of a given query image by first gathering its visual information into a compact descriptor (image representation), then matching it against a database of references with known geolocations. This task can be extremely challenging due to short term appearance changes (e.g., illumination, occlusion and weather) as well as long term variations (e.g., seasonal changes, construction and vegetation). Therefore, a robust VPR technique should be capable of producing descriptors that are invariant to these changes.