← Back to Videos

VINO Learns 3D Scene Features without Labels

Loading video

VINO replaces a hierarchical point backbone with a Volume Transformer and redesigns its training objectives to learn semantic 3D features from unlabeled point clouds.

3D Vision Self-Supervised Learning Point clouds Transformer Scene Understanding

Category: perception

Author: @robotsdigest

Date: 2026-11-07T00:00:00

Duration: 13.8s

Source: https://x.com/robotsdigest/status/2119048844067667991

SSIO: Self-Supervised Inertial Odometry from Raw IMU

SSIO: Self-Supervised Inertial Odometry from Raw IMU

@robotsdigest · 114 views · 2026-10-29

SSIOInertial Odometry

SSIO: Self-Supervised Inertial Odometry Without Labels

SSIO: Self-Supervised Inertial Odometry Without Labels

@QiuYuhengQiu · 128 views · 2026-10-28

SSIOInertial Odometry

ABot-Recon Streams 3D Reconstruction over Long Video

ABot-Recon Streams 3D Reconstruction over Long Video

@cutetoxicguy · 17 views · 2026-11-07

3D reconstructionSLAM

VoxLution Generates 3D Scenes From Sparse Real Scans

VoxLution Generates 3D Scenes From Sparse Real Scans

@angelaqdai · 117 views · 2026-11-02

3D GenerationScene Reconstruction

DeFM: A Depth Foundation Model for Robotics

DeFM: A Depth Foundation Model for Robotics

@robotsdigest · 139 views · 2026-10-28

Depth perceptionFoundation Models

ROS 2 3D LiDAR Localization on Prior Maps

ROS 2 3D LiDAR Localization on Prior Maps

@rsasaki0109 · 191 views · 2026-10-16

ROS2LiDAR Localization

As an Amazon Associate, we earn from qualifying purchases.