На этом MM-SpatialAI-воркшопе были интересные постеры:
- Efficient Feature-Free Initialization for Monocular Visual-Inertial Systems Using a Feed-Forward 3D Model (проект пока без кода https://github.com/Yuantai-Z/FF-VIO-Init)
- FlowHOI: Flow-based Semantics-Grounded Generation of Hand-Object Interactions for Dexterous Robot Manipulation (проект с кодом https://huajian-zeng.github.io/projects/flowhoi/)
-RADIO-ViPE: Online Tightly Coupled Multi-Modal Fusion for Open-Vocabulary Semantic SLAM in Dynamic Environments (проект с кодом https://be2rlab.github.io/radio_vipe)
- DCReg: Decoupled Characterization for Efficient Degenerate LiDAR Registration (код https://github.com/JokerJohn/DCReg)
- ScaRF-SLAM: A Framework for Globally Consistent Online Visual Dense Reconstruction (код https://github.com/ori-drs/ScaRF-SLAM)
- Learning to Localize Reference Trajectories in Image-Space for Visual Navigation (проект с кодом https://finnbusch.com/lotis)
- ScanNet-SG: A Large-Scale Dataset for 3D Scene Graph Alignment (код и данные https://github.com/tud-amr/ScanNet-SG)
- UniFField: A Generalizable Unified Neural Feature Field for Visual, Semantic, and Spatial Uncertainties in Any Scene (проект пока без кода https://sites.google.com/view/uniffield)