ChaLearn
ChaLearn
  1. Home
  2. Long-Term Thermal Imaging (LTD, ECCV'22)
  3. Description

Long-Term Thermal Imaging (LTD, ECCV'22)

Dataset description

The LTD dataset used in the Seasons in Drift Challenge at ECCV'22 is an extension of an existing concept drift dataset and spans 188 days in the period of 14th May 2020  to 30th of April 2021, with a total of 1689 2-minute clips sampled at 1fps with associated bounding box annotations for 4 classes (Human, Bicycle, Motorcycle, Vehicle). The collection of this dataset has included data from all hours of the day in a wide array of weather conditions overlooking the harborfront of Aalborg, Denmark. In this dataset depicts the drastic changes of appearance of the objects of interest as well as the scene over time in a static surveillance context to develop robust algorithms for real-world deployment.

Camera Setup    Camera View 

Camera Setup                                                                                  

Statistics

######### Object Size Grouping Scheme #########
Small (<1024 pixels)
Medium (1025-9695 pixels)
Large (>9696 pixels)

############### Subset Overview ###############
Subset name   :  Full-All
Clips         :  1689
Different days:  188
Timespan      :  2020-05-14 - 2021-04-30
-------------- Object Presence ----------------
Empty frames  : 844638 (78.9937217499792 %)
Frames /w obj : 224609 (21.00627825002081 %)
Total frames  : 1069247

########### Object Distributions ##############
All*          : 6868067
bicycle       : 293280
human         : 5841139
motorcycle    : 32393
vehicle       : 701255
Unique Objects: 143294
------------------- Small ---------------------
All*          : 6092590
bicycle       : 288081
human         : 5663804
motorcycle    : 27153
vehicle       : 113552
------------------- Medium --------------------
All*          : 37468
bicycle       : 7
human         : 454
vehicle       : 37007
------------------- Large ---------------------
All*          : 738009
bicycle       : 5192
human         : 176881
motorcycle    : 5240
vehicle       : 550696
###############################################

Links for Download and Challenge Instructions

As mentioned in the challenge description and challenge rules, each track has an associated train set, defined below.

  • Track 1: Detection at day level (competition link): Train on a predefined and single day data and evaluate concept drift across time. The day is the 13th of February 2020 as it is the coldest day in the recorded data, due to the relative thermal appearance of objects being the least varied in colder environments this is our starting point.
  • Track 2: Detection at week level (competition link): Train on a predefined and single week data and evaluate concept drift across time. The week selected is the week of the 13th – 20th of February 2020  - (i.e. expanding from our starting point)
  • Track 3: Detection at month level (competition link): Train on a predefined and single month data and evaluate concept drift across time. And the selected month is the entire month of February.

The data is split in this manner so as to have an anchor point in the temporal domain from where we measure the concept drift from.

By downloading the data, you agree with the Terms and Conditions of the Challenge. All files are encrypted! To discompress the data, use the associated keys. Decryption keys are provided on Codalab after registration, based on the schedule of the challenge.

Decompressing the data: on Ubuntu, you can install 7zip and decompress the train data as follows:

$ sudo apt-get update

$ sudo apt-get install p7zip-full

$ 7z x Train.zip.001

Valid.zip.006, Valid.zip.007, Valid.zip.008, Valid.zip.009, Valid.zip.010, Valid.zip.011, Valid.zip.012,

Valid.zip.013, Valid.zip.014, Valid.zip.015, Valid.zip.016, Valid.zip.017

News

There are no news registered in Long-Term Thermal Imaging (LTD, ECCV'22)