News

Recent developments

  • 2021-03-19 | We released the FMO deblurring benchmark.
  • 2021-03-19 | DeFMO has been accepted to CVPR 2021! [GitHub]
  • 2020-06-30 | Deblatting has been implemented in Python and PyTorch [GitHub].
  • 2020-04-13 | Android mobile application [app.apk] is released for testing.
  • 2020-02-23 | Paper “Sub-frame Appearance and 6D Pose Estimation of Fast Moving Objects” accepted for CVPR 2020. Data and code are coming soon.
  • 2019-09-16 | We received Honorable Mention at GCPR 2019 for “Non-Causal Tracking by Deblatting”.
  • 2019-09-08 | Matlab implementation of TbD(-NC) is released at GitHub. TbD dataset is updated.
  • 2019-09-02 | Two papers accepted at VOT2019 Workshop at ICCV2019 and at GCPR2019 as oral.
  • 2019-08-01 | Code released [GitHub] for Real-Time Fast Moving Objects Detection.
  • 2019-07-09 | Website fixed, added TbD paper, MSc thesis, TbD results.
  • 2019-05-23 | TbD dataset released.

Sequences

Video files with ground truth

Falling objects dataset

Falling objects dataset as a set of PNG images both for high-speed (ground truth) and low-speed versions is available here (together with ground truth annotations).

TbD-3D dataset

TbD-3D dataset with ground truth in Matlab format can be downloaded here. Dataset as a set of PNG images both for high-speed (ground truth) and low-speed versions is available here.

TbD dataset

TbD with ground truth can be downloaded here, and as PNG images here. Results of the TbD method on the TbD dataset here.

FMO dataset

The whole FMOv2 dataset (decomposed to png images) is available here.

volleyball1 volleyball_passing darts1 darts_window1 softball william_tell tennis_serve_side tennis_serve_back tennis1 hockey squash frisbee blue ping_pong_paint ping_pong_side ping_pong_top

FMOv2 extension

tennis2 more_balls atp_serves

Ground truth

Separate ground truth is available the FMO data set (MATLAB, text) and the FMOv2 data set (MATLAB, text). Ground truth in png format for the whole FMOv2 dataset is here.

Specification of the used text format: text formats.

Videos found on YouTube with fast moving objects

Dodgeball (27M views): https://youtu.be/Spu6OlAZHUo

Beach volleyball (17K views): https://youtu.be/1_ObVLMZS-0

Air hockey (24K views): https://youtu.be/BchGttA8k-Q

Security camera (1M views): https://youtu.be/CkVJyAKwByw, events: Bird [0:09], Head [1:27], Rabbit [2:40], Hand [3:52], Ceiling fan [5:03]

Trick shots (209M views): https://youtu.be/A2FsgKoGD04, events: Cap [1:01], Keys [1:49], Toast [3:34], Newspaper [3:39], Tennis ball [4:05]

Falling watermelon (15M views): https://youtu.be/CkVJyAKwByw, events: first melon [9:22]

Pinball (37K views): https://youtu.be/CkVJyAKwByw, events: start [0:01], bounce [0:21]

Juggling (38K views): https://youtu.be/CkVJyAKwByw, events: first throw [0:06]

Publications

Articles and theses

  • D. Rozumnyi, M. R. Oswald, V. Ferrari, J. Matas, M. Pollefeys, “DeFMO: Deblurring and Shape Recovery of Fast Moving Objects ,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2021. [arxiv,GitHub,Video]
  • J. Kotera, J. Matas, F. Sroubek, “Restoration of Fast Moving Objects,” in IEEE Transactions on Image Processing (TIP), 2020. [Paper]
  • D. Rozumnyi, J. Kotera, F. Sroubek, J. Matas, “Sub-frame Appearance and 6D Pose Estimation of Fast Moving Objects,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2020. [arxiv]
  • J. Kotera, D. Rozumnyi, F. Sroubek, J. Matas, “Intra-frame Object Tracking by Deblatting,” in 7th Visual Object Tracking Challenge Workshop (VOT2019) at ICCV2019. [arxiv]
  • D. Rozumnyi, J. Kotera, F. Sroubek, J. Matas, “Non-Causal Tracking by Deblatting,” in 41th German Conference on Pattern Recognition (GCPR2019). [pdf]
  • D. Rozumnyi, “All-speed Long-term Tracker Exploiting Blur,” MSc Thesis, CTU in Prague, 2019. [pdf]
  • A. Hrabalík, “Implementing and Applying Fast Moving Object Detection on Mobile Devices,” MSc Thesis, FEE, CTU in Prague, 2017. [pdf][extra experiment data]
  • D. Rozumnyi, “Tracking, Learning and Detection over a Large Range of Speeds,” BSc Thesis, FEE, CTU in Prague, 2017. [pdf]
  • D. Rozumnyi, J. Kotera, F. Sroubek, L. Novotny, J. Matas, “The World of Fast Moving Objects,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017. [arxiv]

Applications