Deep learning and computer vision
Mapillary Research is innovating the science behind the world’s leading street‑level imagery platform for understanding places.
About Mapillary Research
Welcome to Mapillary Research, a place where we share scientific papers, data, code, and news! Research at Mapillary covers machine learning using deep learning and computer vision. Two of our main directions are large-scale structure-from-motion and object recognition. These are the fundamental building blocks that power Mapillary and our processing services.
Research and product development are tightly coupled at Mapillary and our results rapidly find their way into production. We publish papers and open source code to share our findings.
Learn more about our team and open positions.
Mapillary Research Team
Meet the people behind the science
Mapillary
Datasets
Train recognition models for street scenes
ECCV 2020 Tutorial
International, City-Scale Computer Vision Benchmarking with Mapillary MegaCities
News
We will organize a tutorial at ECCV 2020 in Glasgow, entitled "International, City-Scale Computer Vision Benchmarking with Mapillary MegaCities"
February, 2020We got 3 papers (1 oral, 2 posters) accepted at CVPR 2020!
February, 2020We are co-organizing the Robust Vision Challenge at ECCV 2020
February, 2020Publications
Learning Multi-Object Tracking and Segmentation from Automatic Annotations
Conf. on Computer Vision and Pattern Recognition (CVPR) 2020 /
Mapillary Street-Level Sequences: A Dataset for Lifelong Place Recognition
Conf. on Computer Vision and Pattern Recognition (CVPR) 2020 /
Modeling the Background for Incremental Learning in Semantic Segmentation
Conf. on Computer Vision and Pattern Recognition (CVPR) 2020 /
The Mapillary Traffic Sign Dataset for Detection and Classification on a Global Scale
arXiv (technical report) /
Towards Generalization Across Depth for Monocular 3D Object Detection
arXiv (technical report) /
Disentangling Monocular 3D Object Detection
International Conf. on Computer Vision (ICCV) 2019 /
Seamless Scene Segmentation
Conf. on Computer Vision and Pattern Recognition (CVPR) 2019 /
Unsupervised Domain Adaptation using Feature-Whitening and Consensus Loss
Conf. on Computer Vision and Pattern Recognition (CVPR) 2019 /
Deep Single Image Camera Calibration with Radial Distortion
Conf. on Computer Vision and Pattern Recognition (CVPR) 2019 /
AdaGraph: Unifying Predictive and Continuous Domain Adaptation through Graphs
Conf. on Computer Vision and Pattern Recognition (CVPR) 2019 /
Boosting Domain Adaptation by Discovering Latent Domains
Conf. on Computer Vision and Pattern Recognition (CVPR) 2018 /
Geometry-Aware Network for Non-Rigid Shape Prediction from a Single View
Conf. on Computer Vision and Pattern Recognition (CVPR) 2018 /
In-Place Activated BatchNorm for Memory-Optimized Training of DNNs
Conf. on Computer Vision and Pattern Recognition (CVPR) 2018 /
AutoDIAL: Automatic DomaIn Alignment Layers
International Conf. on Computer Vision (ICCV) 2017 /
The Mapillary Vistas Dataset for Semantic Understanding of Street Scenes
International Conf. on Computer Vision (ICCV) 2017 /
Loss Max-Pooling for Semantic Image Segmentation
Conf. on Computer Vision and Pattern Recognition (CVPR) 2017 /
Online Learning with Bayesian Classification Trees
Conf. on Computer Vision and Pattern Recognition (CVPR) 2016 /
Dropout Distillation
Intern. Conf. on Machine Learning (ICML) 2016 /