Details of the MIO-TCD dataset

  • The dataset consists of total 786,702 images with 648,959 in the classification dataset and 137,743 in the localization dataset acquired at different times of the day and different periods of the year by thousands of traffic cameras deployed all over Canada and the United States. Those images have been selected to cover a wide range of challenges and are representative of typical visual data captured today in urban traffic scenarios. Each moving object has been carefully identified by nearly 200 persons to enable a quantitative comparison and ranking of various algorithms. This dataset aims to provide a rigorous benchmarking facility for training and testing existing and new algorithms for the classification and localization of moving vehicles in traffic scenes
  • The dataset is divided in two parts : the “classification challenge dataset” and the “localization challenge dataset”.

Classification challenge dataset

  • Contains 648,959 images divided into 11 categories:
    • Articulated truck
    • Bicycle
    • Bus
    • Car
    • Motorcycle
    • Non-motorized vehicle
    • Pedestrian
    • Pickup truck
    • Single unit truck
    • Work van
    • Background

  • The goal of this challenge is to correctly label each image
  • Click here to download the python code for computing training error metrics.

Localization challenge dataset

  • Contains 137,743 high-resolution images containing one (or more) foreground object(s) with one of the following 11 labels:
    • Articulated truck
    • Bicycle
    • Bus
    • Car
    • Motorcycle
    • Motorized vehicle (i.e. Vehicles that are too small to be labeled into a specific category)
    • Non-motorized vehicle
    • Pedestrian
    • Pickup truck
    • Single unit truck
    • Work van

  • The goal of this challenge is to correctly localize and classify each foreground object.
  • Click here to download the python code for computing training error metrics as well as loading and saving bounding boxes.

Click here to download the entire dataset: MIO-TCD-Classification and MIO-TCD-Localization

  • Creative Commons License

    This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

  • Participants are free to upload results for the classification AND/OR localization challenge.

    Click on the tabs below to view image samples.

    classification

    localization