Code for the CVPR2021 paper "Patch-NetVLAD: Multi-Scale Fusion of Locally-Global Descriptors for Place Recognition"

[ECCV-2020 (spotlight)] Self-supervising Fine-grained Region Similarities for Large-scale Image Localization. 🌏 PyTorch open-source toolbox for image-based localization (place recognition).

Official code for CVPR 2022 (Oral) paper "Deep Visual Geo-localization Benchmark"

Code for the RA-L (IROS) 2021 paper "A Hierarchical Dual Model of Environment- and Place-Specific Utility for Visual Place Recognition"

a deep accent recognition network

TensorFlow Implementation of "Learnable Pooling Methods for Video Classification".

NetVLAD based weakly supervised learning whole image descriptor with street view data

[IROS2021] NYU-VPR: Long-Term Visual Place Recognition Benchmark with View Direction and Data Anonymization Influences

Rewrite the LOUPE library (https://github.com/antoine77340/LOUPE) into Keras version. Many learnable pooling or differentiable aggregation methods are covered. (NetVLAD, NetRVLAD, SoftDBoW, NetFV, CG)

This is a Keras implementation of the Netvlad network for visual place recognition.

This repository contains the code for our papers "Learning Condition Invariant Features for Retrieval-Based Localization from 1M Images" and "Geometrically Mappable Image Features".

The 19th Place Solution to The 2nd YouTube-8M Video Understanding Challenge

TensorFlow implementation of end-to-end trainable Triangulation Embedding.

Sejong Building Retrieval

Place recognition using Bag of Visual Words

Make Rendered Database using Rendered Image Obtained by Trained Gaussian Splatting

PyTorch Classification

Extend CosPlace for Visual Geo-localization research, exploring architectures, data augmentation, and optimization for improved performance.

Deep Learning Course Assignment on CNNs and NetVLAD