README.md

metadata

license: gpl-3.0
tags:
  - Tracking
  - VOT
  - Video
  - Motion
  - Propagation
pretty_name: >-
  TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the
  Wild
size_categories:
  - 10K<n<100K

TrackingNet devkit

This repository contains the data of the paper TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild.

Download from HuggingFace

Download splits

from huggingface_hub import snapshot_download

# Download TRAIN_0 split (90GB)
snapshot_download(repo_id="SilvioGiancola/TrackingNet", 
                  repo_type="dataset", revision="main",
                  local_dir="TrackingNet_HF", 
                  allow_patterns="*TRAIN_0/*")
# Download TEST split (35GB)
snapshot_download(repo_id="SilvioGiancola/TrackingNet", 
                  repo_type="dataset", revision="main",
                  local_dir="TrackingNet_HF", 
                  allow_patterns="*TEST/*")
# Download all TRAIN splits (1.2TB)
snapshot_download(repo_id="SilvioGiancola/TrackingNet", 
                  repo_type="dataset", revision="main",
                  local_dir="TrackingNet_HF", 
                  allow_patterns="*TRAIN*")

TrackingNet pip package

conda create -n TrackingNet python pip
pip install TrackingNet

Utility functions for TrackingNet

from TrackingNet.utils import getListSplit

# Get list of codenames for the 12 training + testing split
TrackingNetSplits = getListSplit()
print(getListSplit())
# returns ["TEST", "TRAIN_0", "TRAIN_1", "TRAIN_2", "TRAIN_3", "TRAIN_4", "TRAIN_5", "TRAIN_6", "TRAIN_7", "TRAIN_8", "TRAIN_9", "TRAIN_10", "TRAIN_11"]


# Get list of tracking sequences
print(getListSequence(split=TrackingNetSplits[1])) # return list of tracking sequences in that split
print(getListSequence(split="TEST")) # return list of tracking sequences for testing
print(getListSequence(split=["TRAIN_0", "TRAIN_1"])) # return list of tracking sequences for train splits 0 and 1
print(getListSequence(split="TRAIN")) # return list of tracking sequences for al train splits

Downloading TrackingNet

from TrackingNet.Downloader import TrackingNetDownloader
from TrackingNet.utils import getListSplit

downloader = TrackingNetDownloader(LocalDirectory="path/to/TrackingNet")

for split in getListSplit():
    downloader.downloadSplit(split)