unsupervised-depth-completion-visual-inertial-odometry
simclr
Our great sponsors
unsupervised-depth-completion-visual-inertial-odometry | simclr | |
---|---|---|
2 | 13 | |
183 | 3,927 | |
- | 1.8% | |
5.0 | 2.9 | |
9 months ago | 11 months ago | |
Python | Jupyter Notebook | |
GNU General Public License v3.0 or later | Apache License 2.0 |
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.
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.
simclr
-
Positive and Negative Sampling Strategies for Representation Learning in Semantic Search
For visual representations, you could look into SimCLR and MoCo. https://github.com/google-research/simclr https://github.com/facebookresearch/moco
-
[D] Why is random cropping necessary in SimCLR?
Yeah I think so, it's not hard to check https://github.com/google-research/simclr/blob/2fc637bdd6a723130db91b377ac15151e01e4fc2/data_util.py
-
[R] Deep Learning with a Small Training Batch (or Lack Thereof). Part 1
Code for https://arxiv.org/abs/2006.10029 found: https://github.com/google-research/simclr
-
[D] Current trends in computer vision related to unsupervised learning
SimCLR v2.0 - https://arxiv.org/abs/2006.10029
-
Self-Supervised Contrastive Learning model for video dataset?
My data consists of binary labels (normal & anomalous) where the videos are already broken up into frames in the directory, I'm looking for a model where I can feed a normal-labeled video alongside an anomalous-labeled video like visualized in this example from the SimCLR Repo, a dog will represent the normal video and the chair the anomalous video.
-
[P] Choosing a self-supervised learning framework that's easy to use
No, go to the "tf2" folder in the repo root. https://github.com/google-research/simclr/tree/master/tf2
-
[D] What is meant by width in the SimCLRv2 paper?
Code for https://arxiv.org/abs/2006.10029 found: https://github.com/google-research/simclr
-
[D] Funding PhD in Europe
[1] https://github.com/google-research/simclr [2] https://www.tensorflow.org/tfrc?hl=en&authuser=2
What are some alternatives?
instant-ngp - Instant neural graphics primitives: lightning fast NeRF and more
swav - PyTorch implementation of SwAV https//arxiv.org/abs/2006.09882
dino - PyTorch code for Vision Transformers training with the Self-Supervised learning method DINO
Unsupervised-Classification - SCAN: Learning to Classify Images without Labels, incl. SimCLR. [ECCV 2020]
calibrated-backprojection-network - PyTorch Implementation of Unsupervised Depth Completion with Calibrated Backprojection Layers (ORAL, ICCV 2021)
SimCLR - PyTorch implementation of SimCLR: A Simple Framework for Contrastive Learning of Visual Representations
xivo - X Inertial-aided Visual Odometry
torchlars - A LARS implementation in PyTorch
void-dataset - Visual Odometry with Inertial and Depth (VOID) dataset
Supervised-Constrastive-Learning-in-TensorFlow-2 - Implements the ideas presented in https://arxiv.org/pdf/2004.11362v1.pdf by Khosla et al. [Moved to: https://github.com/sayakpaul/Supervised-Contrastive-Learning-in-TensorFlow-2]
bpycv - Computer vision utils for Blender (generate instance annoatation, depth and 6D pose by one line code)
CodeSearchNet - Datasets, tools, and benchmarks for representation learning of code.