manydepth
unsupervised-depth-completion-visual-inertial-odometry
manydepth | unsupervised-depth-completion-visual-inertial-odometry | |
---|---|---|
2 | 2 | |
596 | 185 | |
1.0% | - | |
0.0 | 5.0 | |
9 months ago | 10 months ago | |
Python | Python | |
GNU General Public License v3.0 or later | GNU General Public License v3.0 or later |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
manydepth
-
How Many Sensors for Autonomous Driving?
Yup! This kind of reconstruction is known as multi-view reconstruction. Though the cameras don't need to have a movable mount, they're already on a car which moves! The car moves and gives them a new "perspective" at every frame. That's how some monocular systems already work. Here's an example of one such system: https://github.com/nianticlabs/manydepth
That said, I think what you're referring to is more extreme perspectives that shift in ways the car cannot drive and you are correct that this would aid in reconstruction. This is how NERF models do their 3D reconstruction (https://nerfies.github.io/).
-
The Temporal Opportunist: Self-Supervised Multi-Frame Monocular Depth – method outperforms all self-supervised methods on KITTI and Cityscapes
Looks like it's available now.
unsupervised-depth-completion-visual-inertial-odometry
-
Unsupervised Depth Completion from Visual Inertial Odometry
Hey there, interested in camera and range sensor fusion for point cloud (depth) completion?
Here is an extended version of our [talk](https://www.youtube.com/watch?v=oBCKO4TH5y0) at ICRA 2020 where we do a step by step walkthrough of our paper Unsupervised Depth Completion from Visual Inertial Odometry (joint work with Fei Xiaohan, Stephanie Tsuei, and Stefano Soatto).
In this talk, we present an unsupervised method (no need for human supervision/annotations) for learning to recover dense point clouds from images, captured by cameras, and sparse point clouds, produced by lidar or tracked by visual inertial odometry (VIO) systems. To illustrate what I mean, here is an [example](https://github.com/alexklwong/unsupervised-depth-completion-visual-inertial-odometry/blob/master/figures/void_teaser.gif?raw=true) of the point clouds produced by our method.
Our method is light-weight (so you can run it on your computer!) and is built on top of [XIVO] (https://github.com/ucla-vision/xivo) our VIO system.
For those interested here are links to the [paper](https://arxiv.org/pdf/1905.08616.pdf), [code](https://github.com/alexklwong/unsupervised-depth-completion-visual-inertial-odometry) and the [dataset](https://github.com/alexklwong/void-dataset) we collected.
-
[N][R] ICRA 2020 extended talk for Unsupervised Depth Completion from Visual Inertial Odometry
In this talk, we present an unsupervised method (no need for human supervision/annotations) for learning to recover dense point clouds from images, captured by cameras, and sparse point clouds, produced by lidar or tracked by visual inertial odometry (VIO) systems. To illustrate what I mean, you can visit our github page for examples (gifs) of point clouds produced by our method.
What are some alternatives?
VIBE - Official implementation of CVPR2020 paper "VIBE: Video Inference for Human Body Pose and Shape Estimation"
instant-ngp - Instant neural graphics primitives: lightning fast NeRF and more
calibrated-backprojection-network - PyTorch Implementation of Unsupervised Depth Completion with Calibrated Backprojection Layers (ORAL, ICCV 2021)
dino - PyTorch code for Vision Transformers training with the Self-Supervised learning method DINO
involution - [CVPR 2021] Involution: Inverting the Inherence of Convolution for Visual Recognition, a brand new neural operator
STEPS - This is the official repository for ICRA-2023 paper "STEPS: Joint Self-supervised Nighttime Image Enhancement and Depth Estimation"
xivo - X Inertial-aided Visual Odometry
second.pytorch - SECOND for KITTI/NuScenes object detection
simclr - SimCLRv2 - Big Self-Supervised Models are Strong Semi-Supervised Learners
void-dataset - Visual Odometry with Inertial and Depth (VOID) dataset
bpycv - Computer vision utils for Blender (generate instance annoatation, depth and 6D pose by one line code)