The ten appendices collect the reference material that the chapters lean on but do not stop to derive. Reach for them the way you reach for a datasheet: when you need the exact definition of a wavelet, the pinout of an inertial measurement unit, the download URL for a benchmark, or the precise formula behind a metric you are about to report. Each appendix is self-contained, so you can jump straight to the one you need without reading the others first.
Appendices A and B refresh the mathematics and deep-learning background. Appendices C through E cover the physical and software substrate: sensors, tools, and data. Appendices F and G handle measurement and simulation. Appendices H through J support teaching and self-study with syllabi, worked solutions, and a unified notation table.
The Ten Appendices
What Each Appendix Covers
| Appendix | Title | Reach for it when you need |
|---|---|---|
| A | Mathematical and Signal-Processing Reference | Linear algebra, probability, Fourier and wavelet transforms, filtering, and sampling identities stated compactly for quick lookup, for example the discrete Fourier transform \(X_k = \sum_{n=0}^{N-1} x_n\, e^{-i 2\pi k n / N}\). |
| B | Deep Learning Refresher for Sequences and Tensors | A concise recap of convolutions, recurrent and attention layers, normalization, and optimization, framed around the tensor shapes that sensor streams produce. |
| C | Sensor Hardware Guide | Practical specifications for accelerometers, gyroscopes, microphones, cameras, radar, and lidar, with interfaces, sampling rates, and noise characteristics. |
| D | The Sensory AI Toolchain | The libraries and frameworks used throughout the book for acquisition, preprocessing, modeling, and deployment, with install notes and version guidance. |
| E | Sensor Datasets and Benchmarks | A catalog of public datasets and benchmarks across modalities, with sizes, licenses, tasks, and download pointers for reproducing chapter experiments. |
| F | Evaluation Metrics Reference | Definitions and formulas for classification, detection, regression, and time-series metrics, with guidance on when each applies and how to avoid common misreports. |
| G | Synthetic Data and Simulation Resources | Simulators, physics engines, and generative pipelines for synthesizing labeled sensor data when real recordings are scarce or expensive. |
| H | Course Syllabi | Ready-to-adapt schedules that map the book onto a one-semester or two-semester course, with pacing, prerequisites, and assessment suggestions. |
| I | Solutions to Selected Exercises | Worked solutions and answer sketches for a representative subset of end-of-chapter exercises, useful for self-checking and for instructors. |
| J | Notation and Glossary | The unified symbol table and a glossary of terms, so that a symbol such as \(\mathbf{x}_t\) means the same thing in every chapter. |