All Projects → shawLyu → Hr Depth

shawLyu / Hr Depth

Licence: mit
[AAAI 2021] HR-Depth : High Resolution Self-Supervised Depth Estimation

Projects that are alternatives of or similar to Hr Depth

Kg Beijing
北京知识图谱学习小组
Stars: ✭ 1,554 (+1239.66%)
Mutual labels:  jupyter-notebook
Lnpr book codes
Codes for Lecture Notes in Probabilistic Robotics
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Jwst
Python library for science observations from the James Webb Space Telescope
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Coursera reinforcement learning
Coursera Reinforcement Learning Specialization by University of Alberta & Alberta Machine Intelligence Institute
Stars: ✭ 114 (-1.72%)
Mutual labels:  jupyter-notebook
Robust Physical Attack
Physical adversarial attack for fooling the Faster R-CNN object detector
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Baselines Results
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Machine Learning Demystified
A weekly workshop series at ITP to teach machine learning with a focus on deep learning
Stars: ✭ 114 (-1.72%)
Mutual labels:  jupyter-notebook
Nlp Models Tensorflow
Gathers machine learning and Tensorflow deep learning models for NLP problems, 1.13 < Tensorflow < 2.0
Stars: ✭ 1,603 (+1281.9%)
Mutual labels:  jupyter-notebook
Text generators
Python code for building a text generator using LSTMs.
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Cpm
Convolutional Pose Machines in TensorFlow
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Ml19 20a
CS 771A: Introduction to Machine Learning, IIT Kanpur, 2019-20-autumn offering
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Gutenberg Poetry Corpus
A corpus of poetry from Project Gutenberg
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Traffic Signs
Building a CNN based traffic signs classifier.
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Quantum Algorithms Tutorials
Tutorials for Quantum Algorithms with Qiskit implementations.
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Flotilla
Reproducible machine learning analysis of gene expression and alternative splicing data
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Cadl
ARCHIVED: Contains historical course materials/Homework materials for the FREE MOOC course on "Creative Applications of Deep Learning w/ Tensorflow" #CADL
Stars: ✭ 1,478 (+1174.14%)
Mutual labels:  jupyter-notebook
A Nice Mc
Code for "A-NICE-MC: Adversarial Training for MCMC"
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Emp
Code repository of the Earth Microbiome Project.
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook
Dat8
General Assembly's 2015 Data Science course in Washington, DC
Stars: ✭ 1,516 (+1206.9%)
Mutual labels:  jupyter-notebook
Bayareadlschool
Slides and exercises for the Theano tutorial at the Deep Learning School in Stanford, September 24-25, 2016
Stars: ✭ 115 (-0.86%)
Mutual labels:  jupyter-notebook

HR-Depth: High Resolution Self-Supervised Monocular Depth Estimation

This is the official implementation for training and testing depth estimation using the model proposed in

HR-Depth: High Resolution Self-Supervised Monocular Depth Estimation

Xiaoyang Lyu, Liang Liu, Mengmeng Wang, Xin Kong, Lina Liu, Yong Liu*, Xinxin Chen and Yi Yuan.

This paper has been accepted by AAAI 2021.

Qualitative_Result

Note: We temporarily release the evaluation version and some pretrained models of our paper. The training codes are modified according to Monodepth2, and we will release them soon.

Update

2021.1.27

  1. The training code will be released around the beginning of the March.
  2. For re-implementing HR-Depth, you can clone Monodepth2 and simply replace the DepthDecoder with HRDepthDecoder. Our parameter settings are exactly the same as Monodepth2.
  3. In our paper, we wrote the initial learning rate wrong. It should be 1e-4, not 1e-3. We will fix this mistake in the final version. Thanks for someone pointing out our problem.

Quantitative Results

HR-Depth Results

Quantitative_results_1

Lite-HR-Depth Results

Quantitative_result_2

Usage

Requirements

Assuming a fresh Anaconda distribution, you can install the dependencies with:

conda install pytorch=1.5.0 torchvision=0.6.0 -c pytorch
conda install opencv=4.2
pip install scipy=1.4.1

Pretrained Model

We provided pretrained model as follow:

Model Name Resolution Dataset Supervision Abs_Rel $\delta<1.25$ $\delta<1.25^2$ $\delta<1.25^3$
HR_Depth_CS_K_MS_$640\times192$ $640\times192$ CS+K MS 0.104 0.893 0.964 0.983
HR_Depth_K_MS_$1024\times320$ $1024\times320$ K MS 0.101 0.899 0.966 0.983
HR_Depth_K_M_$1280\times384$ $1280\times384$ K M 0.104 0.894 0.966 0.984
Lite_HR_Depth_K_T_$1280\times384$ $1280\times384$ K T 0.104 0.893 0.967 0.985

KITTI training data

You can download the entire KITTI_raw dataset by running:

wget -i splits/kitti_archives_to_download.txt -P kitti_data/

Then unzip with

cd kitti_data
unzip "*.zip"
cd ..

Warning: The size of this dataset is about 175GB, so make sure you have enough space to unzip them.

KITTI evaluation

--data_path: path of KITTI dataset --load_weights_folder: path of models --HR_Depth: inference by HR-Depth --Lite_HR_Depth: inference by Lite-HR-Depth

To prepare the ground truth depth maps run:

python export_gt_depth.py --data_path ./kitti_RAW

assuming that you have placed the KITTI RAW dataset in the default location of ./kitti_data.

For HR-Depth:

python evaluate_depth.py --data_path ./kitti_RAW --load_weights_folder ./models/HR_Depth_CS_K_MS_640x192 --HR_Depth

python evaluate_depth.py --data_path ./kitti_RAW --load_weights_folder ./models/HR_Depth_K_M_1280x384 --HR_Depth 

For Lite-HR-Depth:

python evaluate_depth.py --data_path ./kitti_RAW --load_weights_folder ./models/Lite_HR_Depth_K_T_1280x384 --Lite_HR_Depth
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].