README

Updated: 4th October 2018.

Computer Vision and Intelligence Group, IIT Madras

This repository contains content that we use for CNN SLAM. Original paper

Contributors

Heterogenous graphs; convert every CPU op into a compute graph and note performance with placement.
Stereo matching methods and GPU optimisation.
CPU efficient object detectors and depth estimators.
Indoor optimised depth estimation (monocular.)

Name		Name	Last commit message	Last commit date
Latest commit History 100 Commits
FCRN		FCRN
Papers		Papers
bundle_adjustment		bundle_adjustment
camera_calib		camera_calib
dummy_data		dummy_data
g2opy		g2opy
g3docs		g3docs
graph_optimization		graph_optimization
monodepth		monodepth
monodepth_infer		monodepth_infer
pangolin		pangolin
pose_estimation		pose_estimation
pose_graph_optimisation		pose_graph_optimisation
rgbd_ptam		rgbd_ptam
stereo_ptam		stereo_ptam
utils		utils
.gitignore		.gitignore
README.md		README.md
Run_depth_engine.py		Run_depth_engine.py
To do		To do
avatar.png		avatar.png
core		core
keyframe_utils.py		keyframe_utils.py
params.py		params.py
run.py		run.py