profile
viewpoint
Prakhar Kulshreshtha ankuPRK Carnegie Mellon University Pittsburgh, USA https://ankuprk.github.io/ Pursuing MS in Computer Vision (MSCV) at CMU

dhishku/Machine-Learning-for-Grain-Assaying 11

This project aims at creating mobile based solutions for scientific grading/assaying of food grains. Aim is to be used in e-National Agriculture Market (NAM)

ankuPRK/Emotion-Recognition-in-Hindi-Speech 4

Classifying utterances in Hindi speech in one of the 8 emotional states (anger, fear, disgust, neutral, sad, happy, surprise, sarcastic) in spoken speech in Hindi

ankuPRK/COFC 3

An Online Algorithm for Constrained Face Clustering in Videos

ankuPRK/DIP-Grain-Project 1

Course project for EE602A

ankuPRK/Face-Detection-Tracking-and-Clustering 1

We detect and track faces in video, then extract features from those face tracks and try to cluster them into given number of persons.

anandbhoraskar/argoverse-forecasting 0

Official Repository for Argoverse Motion Forecasting Baselines

anandbhoraskar/ORB_SLAM3 0

ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM

ankuPRK/academic-kickstart 0

a fork of academic-kickstart project

ankuPRK/ankuPRK.github.io 0

homepage repository

issue commentlifelong-robotic-vision/lifelong-slam

Understanding trans_matrix.yaml

Hi @cedrusx sorry for forgetting to close the issue! Yes your response on Sep 12 makes sense. Thank you! :)

ankuPRK

comment created time in 2 days

push eventankuPRK/LDU_Factorization

Prakhar Kulshreshtha

commit sha 08e1144fd0af01b7aac5e7c6c9c5bc8c18da3e72

Update README.md

view details

push time in 16 days

push eventankuPRK/LDU_Factorization

Prakhar Kulshreshtha

commit sha 064b7652673376d61c3de44dd1acb264f3e22e8d

Update README.md

view details

push time in 16 days

push eventankuPRK/LDU_Factorization

Prakhar Kulshreshtha

commit sha 0e194f4662545e7d53b8792ca9a304bb4b44f0fc

Update README.md

view details

push time in 16 days

push eventankuPRK/LDU_Factorization

Prakhar Kulshreshtha

commit sha bfed86fa5b602e21a79d906f5a1fe2490151bf29

Create README.md

view details

push time in 16 days

push eventankuPRK/LDU_Factorization

ankuPRK

commit sha 1028a5cca339184c8c9d56f5b5a3c84faad24d86

cleared cache

view details

push time in 16 days

push eventankuPRK/LDU_Factorization

ankuPRK

commit sha 685729d8c2039d4945b76c775200f001714ef7e4

added the code

view details

push time in 16 days

create barnchankuPRK/LDU_Factorization

branch : main

created branch time in 16 days

created repositoryankuPRK/LDU_Factorization

Factorizing a rectangular matrix A as PA=LDU

created time in 16 days

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha c09814c47195e62b4ebc45cd7cd1760291208e53

added homography image

view details

push time in 16 days

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha eb81cc770e9ef4c18dcd4bcea7b8e8906fdfd7b7

corrected date in resume

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha bc55e7a1199d69413392fe96395e90633de4d647

added more details in resume

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha e6453922996d6fe053222f6b425e4ebdbdcdfb54

updated resume

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha 805cadd6676c7073220c6fa38c6d09a631f9866c

updated resume

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha dee2f24dea1ab089ef5a4b0a12a3e2ac25594c28

updated the home pic

view details

push time in a month

issue commentUZ-SLAMLab/ORB_SLAM3

Question about relocalization in tracking.cc

welcome :)

xieqi1

comment created time in a month

issue commentUZ-SLAMLab/ORB_SLAM3

Question about relocalization in tracking.cc

I think they have replaced relocalization with submap merging i.e. every new sequence will start with a new sub-map, which will then be aligned and merged with the one of the submaps in the set of maps we already have, given it meets certain criteria.

xieqi1

comment created time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha 2f94bd95c749c98a1147f2c75bc44878da88c434

updated resume 2

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha 1399bd9690258a9cc00968454c4fa24d087872e7

updated resume

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha a5f6bf50f97210f96c1d22a0354306377d5e1105

added capstone project

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha aa18b23162b7c69e34de0be42df57062fcfb23fd

updated publication figure

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha 9d6b27d803d7e911cb82a05440e8d29bf9c46d36

added a publication

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha 600ee9953bf4ca57b26f7c4a98b1371d9db06926

updated intro and resume

view details

push time in a month

push eventankuPRK/ankuPRK.github.io

ankuPRK

commit sha 04608fc03555be918bf4af5ed17d489e92e7f9f1

updated the intro

view details

push time in a month

issue openedUZ-SLAMLab/ORB_SLAM3

RGB-D Camera.bf is not getting set

https://github.com/UZ-SLAMLab/ORB_SLAM3/blob/ef9784101fbd28506b52f233315541ef8ba7af57/src/Tracking.cc#L622

Hello, I think we need to replace this: if(mSensor==System::STEREO || mSensor==System::IMU_STEREO) with: if(mSensor==System::STEREO || mSensor==System::IMU_STEREO || mSensor==System::RGBD)

Otherwise mbf stays 0 which makes mThDepth = 0 for RGBD (i.e. all points will be considered far points) https://github.com/UZ-SLAMLab/ORB_SLAM3/blob/ef9784101fbd28506b52f233315541ef8ba7af57/src/Tracking.cc#L663

created time in a month

push eventanandbhoraskar/ORB_SLAM3

ankuPRK

commit sha 48b9d5e817b717eea2a7ec44f7248a124f2c0575

fisheye stereo yaml for openloris

view details

push time in a month

push eventanandbhoraskar/ORB_SLAM3

ankuPRK

commit sha e0c0461127a714ef701934809307a00e462cf850

support for sequences in RGB-D

view details

push time in 2 months

push eventanandbhoraskar/ORB_SLAM3

ankuPRK

commit sha ebbad261557b82e6057dd7468f52847cf2d2db1f

config file for stereo fisheye loris

view details

push time in 2 months

push eventanandbhoraskar/ORB_SLAM3

ankuPRK

commit sha 5e60e6bd13cdba953122ebcdc0a0cfed5b88317a

added code for trajectory evaluation

view details

push time in 2 months

issue commentUZ-SLAMLab/ORB_SLAM3

Lapping area between images

Thanks again this makes sense! We can keep the issue closed.

ankuPRK

comment created time in 2 months

push eventanandbhoraskar/ORB_SLAM3

ankuPRK

commit sha be748d5c13b6e485c870e78f0dfb75fd547ec67c

added Loris parsing functionalities as jnb (part 1/2)

view details

push time in 2 months

issue commentUZ-SLAMLab/ORB_SLAM3

Lapping area between images

Thank`you! The explanation makes perfect sense. Just one more follow-up doubt. In the yaml file lapping.left=0 and lapping.right=511, so is the origin of these coordinates center of left camera?

ankuPRK

comment created time in 2 months

issue commentUZ-SLAMLab/ORB_SLAM3

Lapping area between images

I don't know I tried looking it up but couldn't find anything. That's why I got confused.

ankuPRK

comment created time in 2 months

issue openedUZ-SLAMLab/ORB_SLAM3

Lapping area between images

Could someone help about what is the lapping area for fisheye stereo config file TUM_512.yaml?

# Lapping area between images
Lapping.left: 0
Lapping.right: 511

I was working on a dataset which uses t265 fisheye stereo camera so I wanted to know how to set this parameter. Thanks!

created time in 2 months

issue closedUZ-SLAMLab/ORB_SLAM3

Extending ORBSLAM3 to RGBD+Inertial

Hey guys I have some data where I have RGBD data along with IMU measurements.

I was wondering if any feasibility issue will arise if I try to merge TUM RGDB system with EUROC stereo-imu such that the depth representation of TUM replaces Stereo of EUROC. If I remember correctly in ORBSLAM2 both Stereo and RGBD modalities were internally calling same tracking and mapping utilities.

closed time in 2 months

ankuPRK

issue commentUZ-SLAMLab/ORB_SLAM3

Extending ORBSLAM3 to RGBD+Inertial

Thanks! Sure if I get something meaningful will definitely share! :)

ankuPRK

comment created time in 2 months

issue openedUZ-SLAMLab/ORB_SLAM3

Extending ORBSLAM3 to RGBD+Inertial

Hey guys I have some data where I have RGBD data along with IMU measurements.

I was wondering if any feasibility issue will arise if I try to merge TUM RGDB system with EUROC stereo-imu such that the depth representation of TUM replaces Stereo of EUROC. If I remember correctly in ORBSLAM2 both Stereo and RGBD modalities were internally calling same tracking and mapping utilities.

created time in 2 months

issue openedlifelong-robotic-vision/lifelong-slam

Understanding trans_matrix.yaml

Hey guys could you explain what "base_link" and "marker" mean in trans_matrix.yaml with different sequences.

Also, for a given parent_frame and child_frame, is the extrinsic matrix T_[parent][child] or reverse i.e. if I have coordinates x_[child] then is the extrinsic matrix given here of the form: x_[parent] = T_[parent][child]*x_[child] or vice versa?

created time in 2 months

startedUZ-SLAMLab/ORB_SLAM3

started time in 3 months

more