---
license: apache-2.0
---
Handlabelled Dataset Attributed to Project: DejaView\
AI4CE Lab, NYU Center for Robotics & Embodied Intelligence\
PI: Dr. Chen Feng\
Project Leader: Luca Macesanu\
Dataset annotation pipeline & upload: Greta Perez-Haiek\
And others.

The following locations were scanned then labelled:\
1) 2 metro tech 8 floor\
2) 3rd floor  5 metro tech\
3) 5 metro tech lobby\
4) 10th floor 2 metro tech\
5) 360 Jay street 4 th floor test 2\
6) 360 Jay street 8 th floor\
7) 360 Jay street 10 floor\
8) Outside metro tech 2\

Each location is labelled with a stride length of 5 (so, one image is labelled for every 5 images in both the left/right camera video frame roll).

The following objects were labelled (the same for every location):\
    "Paper_Plate",\
    "Sponge(blue)",\
    "Book",\
    "Pen",\
    "Pencil",\
    "Blue_Tape",\
    "Green_Tape",\
    "Scissors",\
    "Soy_Sauce",\
    "Glove",\
    "Water_Bottle",\
    "Multimeter",\
    "Clamp",\
    "pliers",\
    "Ketchup"

All annotations (x,y) are found within stream_annotations.json in each locations' respective folders: D:\DejaView\[LOCATION]\camera.

questions email... 
Luca Macesanu: lim2045@nyu.edu
Great Perez-Haiek: gp2583@nyu.edu