"FoundationVIO: Enforcing Metric Scale in Feed-forward Visual-Inertial Odometry"
TLDR: 3D foundation models + IMU measurements and factor-graph optimization to recover metric-scale camera trajectories and dense 3D reconstructions, with loop closure for globally consistent SLAM.
🛠️ Software Release 🛠️
PlanarMesh is a real-time algorithm for minimal LiDAR mesh reconstruction:
https://t.co/kBXhTfYWJa
PlanarMesh directly estimates planar surfaces with high fidelity and data compactness - 10x smaller than SotA methods.
The CDT in Autonomous Intelligent Machines & Systems is part of the TechExpert pilot offering stipends of £31,000 for students eligible for home fee status.
Join us and help grow the UK’s national capability in Autonomous Intelligent Machines & Systems.
https://t.co/WOO4n49W1H
The candidate has the opportunity to interact with other groups in ORI as well as our collaborators in Seoul National University, ETH Zurich, Georgia Tech, Hilti and elsewhere on some very exciting projects.
My research group:
https://t.co/urns6OgPQF
🚨 HIRING: New Post-Doc Position on 3D Reconstruction and Mapping!
🕒 Deadline: 27 Nov 2025
📚 Duration: 2 year role
📍 Location: Oxford, UK
Focus areas: visual SLAM, semantic segmentation, NeRF/3DGS but also legged robots
Apply here:
https://t.co/a3tFNHxi9W
Our paper OpenLex3D is accepted to NeurIPS 2025!
tldr; a new benchmark for open-vocabulary scene understanding methods.
Page: https://t.co/eZlwJGZBpI
With thanks to @SachMorin, Martin Büchner, @mmattamala, @kumaraditya_gup, Abhinav Valada, @duckietown_coo & @MauriceFallon
@AjdDavison It is Meta’s slam running on meta’s data from its own hardware. Heavily optimised by maybe 20 people.
It’s really only okvis and openvins that are comparable. Openvins does pretty well from my experiments.
The other systems are mono - so incomparable.
🚨 HIRING: New Post-Doc Position at @oxfordrobots!
I have a new collaborative project working with @DrStephenMellon to further develop our vision-based ultrasound reconstruction system - called ScanLite. Deadline: 8 Sept 2025
Apply here:
https://t.co/68w2vrR2jG
DigiForest researchers from @oxfordrobots demonstrating their mapping technology to students of the Summer School on Forest Ecosystems organised by @m_mokros (@3DForEcoTech ) Sruthi Moorthy (@SalGoTeam)
🚨 HIRING: New Post-Doc Position at @oxfordrobots!
I have a new collaborative project working with @DrStephenMellon to further develop our vision-based ultrasound reconstruction system - called ScanLite. Deadline: 8 Sept 2025
Apply here:
https://t.co/68w2vrR2jG
The focus is vision-based tracking and ultrasound reconstruction for medical applications
Collaborators include @OOEC_Oxford@ndorms as well as clinical partners in @OUHospitals
Details on the project
https://t.co/TePwdZgknT
https://t.co/urns6OgPQF
#slam#3dvision#ultrasound
@gabriberton@wenjing_bian@Yifu_Tao thanks for the input! i guess the tables with our dataset are consistency your own paper - even on challenging night images.
It was fantastic to work with Juan Suarez of @ForestryComm in @digiforest_eu. Our team has been working with European partners to develop robot & sensing tech to digitise forests quickly & accurately.
Papers: https://t.co/yWnlO9PyZ0
Digiforest website: https://t.co/F7RZrQOxPZ
Seeing in the Dark: Benchmarking Egocentric 3D Vision with the Oxford Day-and-Night Dataset
@ziruiwang_, @wenjing_bian, @li_xinghui, @Yifu_Tao, Jianeng Wang, @MauriceFallon, Victor Adrian Prisacariu
tl;dr: egocentric dataset with GT 3D and lighting variation for benchmarking NVS and visual relocalization
https://t.co/jq2KGfVQlz
🚨🚨 New #SLAM#RealityCapture Dataset! 🚨🚨
The Oxford Spires Dataset.
New mapping benchmark in Oxford's famous colleges and Blenheim Palace focusing on high quality reconstruction (#nerf#gaussiansplatting) and colorization.
https://t.co/PGeENpxVeG
#spatialai#oxford
Program involves 1 yr of training at Uni. of Manchester before moving to Oxford, Glasgow, or Manchester for Yr2-4.
Positions will mostly for students from UK and Ireland (🇮🇪!). A fraction can fund internationals students (~30%).
Full details on RAINZ:
https://t.co/MwMpNxZaPt
🚨🚨 Last Reminder: 17 Funded Robotics PhD Positions! 🚨🚨
In the new RAINZ PhD training programme. The deadline for application is 11th February 2025.
We have four RAINZ positions at the Oxford Robotics Institute. Details here:
https://t.co/0wYyE5x9qs