At last years competition, multiple teams were utilizing machine learning for their vision processing. Our team used this as well, and after talking to teams at competition and sharing data them, we wanted to open up our years of transdec logs to teams looking for a good place to start on their implementations. Links below will be provided for the past three years of transdec footage, and annotations are included for some of the directiories of images for each year.
The link to the data is:
The data is provided as is, not all of it is useful, but most is. Directories with annotations include the tag _done at the end of the folder name. Folders with _none do not have relevant data. All annotations are in the VOC format. If conversions are necessary, I can write a conversion script to help out, just let me know.
As a note, this data is hosted on my personal NAS. Your are certainly free to download everything, but if you could download it just once I would appreciate it. If you annotate anything else and want to upload it, please let me know and I will get you access to upload. If you have additional images that you cannot share yourself, please ask I can store them for you.
Any other problems or questions feel free to reach out. I cannot guarantee that I can help, but I will do my best or redirect you to someone who can answer. My email is: firstname.lastname@example.org
Hope it helps,
Robotics Association, Embry-Riddle