image_navigation_python's Introduction

Python demo for our CVPR'2019 paper

Mapping, Localization and Path Planning for Image-based Navigation using Visual Features and Map

This code was tested with Python 3.8 and mosek 9.2.29. The easiest way to install mosek is:

conda install -c mosek mosek

Please, execute 'demo.py' to view our demo.

This demo does the following using the concepts introduced in our paper:

Find landmarks in a subsequence of the Oxford Robotcar run from 2015-10-29 12:18:17
Match a short query sequence from 2014-11-18 13:20:12

The precalculated feature distances in this demo are based on features extracted with a VGG-16 + NetVLAD + whitening network. We use the Off-the-shelf on Pitts30k model available on the NetVLAD project page in combination with this NetVLAD TensorFlow implementation.

If you do not have mosek installed, you can have a look at the saved figures in the results folder instead.

The produced outputs are:

Scatter plot of original reference and query sequences
Topology of reference sequence used for finding landmarks with network flow
Selected landmarks
Accuracy vs. distance plot of the final matching

image_navigation_python's People

Contributors

Stargazers

Watchers

image_navigation_python's Issues

Problem with sampling

Hi, Janine

I've tried the code just running demo.py and everything goes fine.
Then, just to check the sampling, I renamed the temp.pickle and rerun the demo.

Then, this error arised:

  File "demo_COLD.py", line 191, in <module>
    main()
  File "demo_COLD.py", line 135, in main
    flow_lm = sample_with_flow(ref['xy'], edges, source_idx, sink_idx, geo_dists, feat_dists, num_landmarks)
  File ..., line 103, in sample_with_flow
    M.constraint(stack, mf.Domain.inRotatedQCone().axis(2))  # Each row is in a rotated quadratic cone
  File "/usr/local/lib/python3.7/dist-packages/mosek/fusion/impl/_implementation.py", line 3554, in constraint
    return self._constraint_Lmosek_4fusion_4Expression_2Lmosek_4fusion_4ConeDomain_2(*args)
  File "/usr/local/lib/python3.7/dist-packages/mosek/fusion/impl/_implementation.py", line 6184, in _constraint_Lmosek_4fusion_4Expression_2Lmosek_4fusion_4ConeDomain_2
    return (self.__constraint_1_SLmosek_4fusion_4Expression_2Lmosek_4fusion_4ConeDomain_2("",_0,_1))
  File "/usr/local/lib/python3.7/dist-packages/mosek/fusion/impl/_implementation.py", line 7668, in __constraint_1_SLmosek_4fusion_4Expression_2Lmosek_4fusion_4ConeDomain_2
    _22=_19[_21]
IndexError: index 2 is out of bounds for axis 0 with size 2

The trouble line was the line 103 in sample_with_flow.py:
M.constraint(stack, mf.Domain.inRotatedQCone().axis(2)) # Each row is in a rotated quadratic cone

Checking the dimensions of the variable stack, its shape is [898588, 3].
So, to solve the problem, I think you have to set:

M.constraint(stack, mf.Domain.inRotatedQCone().axis(**1**)) # Each row is in a rotated quadratic cone

Regards

Recommend Projects

janinethoma / image_navigation_python Goto Github PK

image_navigation_python's Introduction

Mapping, Localization and Path Planning for Image-based Navigation using Visual Features and Map

image_navigation_python's People

Contributors

Stargazers

Watchers

Forkers

image_navigation_python's Issues

Problem with sampling

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent