ENGSCI 304: Engineering Machine Learning II — third-party sources and licences

These notes are provided for non-commercial teaching and learning. Third-party
material retains its own copyright and licence terms; no single licence is
asserted for all material in the book. Chapter captions and references supply
additional credits.

TRACKID3X3 BASKETBALL DATA AND DERIVATIVES

Source: TrackID3x3, Indoor subset, basket_S1T4_pre, selected frames and supplied
court landmarks. Creators: Kazuhiro Yamada, Li Yin, Qingrui Hu, Ning Ding,
Shunsuke Iwashita, Jun Ichikawa, Kiwamu Kotani, Calvin Yeung and Keisuke Fujii
(2025).
Dataset paper: https://doi.org/10.1145/3728423.3759400
Public archive: https://drive.google.com/drive/folders/1aWqMwQKr5xKMjqms7-raYluSlxPsGvwX
Dataset licence: Creative Commons Attribution 4.0 International (CC BY 4.0).
https://creativecommons.org/licenses/by/4.0/

Course modifications include still extraction, crops, annotation/detection
box overlays, track labels, segmentation-mask overlays, colour-channel views,
geometric overlays and explanatory layouts. The website also includes a short
processed clip of frames 50–100 with SAM3 masks and labels, played in slow
motion. The court helper converts landmark coordinates from centimetres to
metres and supplies a calculated image-to-court transformation. Model outputs
and the explanatory figure styling were created for the course, not provided
as official results by the dataset authors. No endorsement is implied.
The original full video and the practical-notebook package are not distributed
through this website.

FOUNDATIONS OF COMPUTER VISION: PINHOLE IMAGE-FORMATION FIGURE

Copyright 2024 Antonio Torralba, Phillip Isola and William T. Freeman.
Foundations of Computer Vision, MIT Press, Figure 5.3.
Source and caption: https://visionbook.mit.edu/imaging.html#fig-wallpicture
Original image: https://visionbook.mit.edu/figures/imaging/no_picture_on_a_wall_aina.png
Copyright notice: https://visionbook.mit.edu/copyright.html
The book website states Creative Commons Attribution-NonCommercial-NoDerivatives
4.0 International (CC BY-NC-ND 4.0), subject to its exception for components
included with permission. No separate exception is identified in this figure's
caption. https://creativecommons.org/licenses/by-nc-nd/4.0/

The PNG is reproduced unmodified; its course caption is separate. No endorsement
is implied. The dataset's CC BY licence does not apply to this figure. Other
camera diagrams were constructed for these notes; the chapters identify the
book examples that informed the relevant geometric explanations.

OTHER FIGURE SOURCES

The MNIST digit used in the small detector illustrations comes from the MNIST
handwritten-digit database by Yann LeCun, Corinna Cortes and Christopher J. C.
Burges: https://yann.lecun.com/exdb/mnist/
The filtering diagrams and illustrative predictions were created for the course;
they are not published MNIST benchmark results.

The tensor-index diagram is reused from Oliver Maclaren's ENGSCI 205 teaching
material, with surrounding whitespace trimmed. Other explanatory diagrams and
layouts were prepared for ENGSCI 304 unless otherwise credited in the chapters.

SOFTWARE

The rendered site uses Quarto and its bundled web dependencies; their notices
are retained in the distributed files. Referenced model/software packages have
their own licences. Citing or using a package does not transfer the dataset's
licence to that software.