Terms and abbreviations are listed in the order of their first appearance in the chapters.
| Term | Description | Reference |
| LLMs | Large Language Models. Models trained on large text corpora to process and generate language. | Sec. 1.1.1 |
| RL | Reinforcement Learning. Learning to select actions through interaction and reward feedback. | Sec. 1.1.2 |
| GIS | Geographic Information Systems. Systems for storing, managing, and analyzing geographically referenced data. | Sec. 1.1.4 |
| VQA | Visual Question Answering. Answering natural-language questions about visual content. | Sec. 1.1.4 |
| VLN | Vision-and-Language Navigation. Navigation using visual observations and natural-language instructions. | Sec. 1.1.4 |
| GIScience | Geographic Information Science. The study of geographic information, its representation, and its analysis. | Sec. 1.2.2 |
| SLAM | Simultaneous Localization and Mapping. Joint estimation of an agent’s pose and a map of its environment. | Sec. 1.2.2 |
| TSDF | Truncated Signed Distance Function. A signed distance function whose values are limited to a fixed range around a surface. | Sec. 1.2.2 |
| SDF | Signed Distance Function. A function giving the distance to a surface, with a sign distinguishing its two sides. | Sec. 1.2.2 |
| VO | Visual Odometry. Estimation of camera motion from a sequence of images. | Sec. 1.2.3 |
| LiDAR | Light Detection and Ranging. Sensing that uses laser light to measure distances. | Sec. 1.2.3 |
| CT | Computed Tomography. Imaging that reconstructs cross-sectional structure from X-ray measurements. | Sec. 1.2.3 |
| MRI | Magnetic Resonance Imaging. Imaging that uses magnetic fields and radiofrequency signals to depict internal anatomy. | Sec. 1.2.3 |
| CNNs | Convolutional Neural Networks. Neural networks that use convolutional filters to extract local features. | Sec. 1.2.3 |
| BEV | Bird’s-Eye View. A top-down representation of a scene. | Sec. 1.2.3 |
| NeRF | Neural Radiance Field. A continuous representation of scene density and view-dependent radiance for novel-view synthesis. | Sec. 1.2.4 |
| CoT | Chain-of-Thought. A sequence of intermediate reasoning steps used to solve a problem. | Sec. 1.2.4 |
| VLA | Vision-Language-Action Model. A model that maps visual observations and language instructions to actions. | Sec. 1.2.4 |
| VLMs | Vision-Language Models. Models that jointly process visual and linguistic information. | Sec. 1.2.4 |
| 3DGS | 3D Gaussian Splatting. A method for representing and rendering scenes using optimized 3D Gaussians. | Sec. 1.2.4 |
| JEPA | Joint Embedding Predictive Architecture. An architecture that predicts target representations in an embedding space. | Sec. 1.2.5 |
| XR | Extended Reality. An umbrella term for virtual, augmented, and mixed reality. | Sec. 2.2.3 |
| RGB-D | RGB and Depth. Image data combining color channels with depth measurements. | Sec. 4.2 |
| MLP | Multilayer Perceptron. A feedforward neural network with fully connected layers. | Sec. 4.2.6 |
| SfM | Structure from Motion. Recovery of camera poses and 3D structure from overlapping images. | Sec. 4.2.7 |
| DSL | Domain-Specific Language. A programming language designed for a particular class of tasks. | Sec. 5.1 |
| IMU | Inertial Measurement Unit. A sensor unit that measures acceleration and angular velocity. | Sec. 5.5 |
| IoU | Intersection over Union. The size of the overlap between two regions divided by the size of their union. | Sec. 6.1.1 |
| mAP | Mean Average Precision. Average precision averaged across classes or evaluation categories. | Sec. 6.1.1 |
| PnP | Perspective-n-Point. Estimation of camera pose from corresponding 3D points and 2D image locations. | Sec. 6.1.3 |
| MVS | Multi-View Stereo. Dense reconstruction of 3D geometry from images with known camera poses. | Sec. 6.2.1 |
| UAV | Unmanned Aerial Vehicle. An aircraft operated without a pilot on board. | Sec. 6.2.3 |
| RANSAC | Random Sample Consensus. Robust model fitting by testing hypotheses from randomly sampled subsets of observations. | Sec. 6.2.4 |
| LPIPS | Learned Perceptual Image Patch Similarity. A perceptual image distance computed from deep visual features. | Sec. 6.3.5 |
| PSNR | Peak Signal-to-Noise Ratio. An image quality measure derived from mean squared reconstruction error. | Sec. 6.3.5 |
| SSIM | Structural Similarity Index. An image similarity measure based on local luminance, contrast, and structure. | Sec. 6.3.5 |
| VAE | Variational Autoencoder. A generative model that learns a probabilistic latent representation of data. | Sec. 6.5.2 |
| IK | Inverse Kinematics. Finding joint configurations that achieve a desired pose of an end effector or body part. | Sec. 7.1 |
| TAMP | Task and Motion Planning. Joint planning of discrete task actions and feasible continuous motions. | Sec. 7.1.4 |
| MPC | Model Predictive Control. Control that repeatedly optimizes a finite-horizon plan and executes its first action. | Sec. 7.1.4 |
| GNNs | Graph Neural Networks. Neural networks that learn representations by aggregating information over graph connections. | Sec. 7.2.1 |
| POMDP | Partially Observable Markov Decision Process. A sequential decision model in which the agent observes the environment state only indirectly. | Sec. 7.2.5 |
| RAG | Retrieval-Augmented Generation. Generation conditioned on information retrieved from an external collection or memory. | Sec. 7.3.6 |
| ToM | Theory of Mind. Reasoning about the beliefs, intentions, and knowledge of other agents. | Sec. 7.3.7 |
| FVD | Fréchet Video Distance. A distance between feature distributions of generated and reference videos. | Sec. 7.3.8 |
| FID | Fréchet Inception Distance. A distance between Inception feature distributions of generated and reference images. | Sec. 7.3.8 |
| GANs | Generative Adversarial Networks. Generative models trained through competition between a generator and a discriminator. | Sec. 8.1.1 |
| DDIM | Denoising Diffusion Implicit Model. A diffusion sampling formulation that supports deterministic sampling and fewer sampling steps. | Sec. 8.1.3 |
| SDS | Score Distillation Sampling. Optimization of a representation using guidance from a pretrained diffusion model. | Sec. 8.2.1 |
| GNSS | Global Navigation Satellite System. A satellite system providing positioning, navigation, and timing information. | Sec. 9.2.4 |
| BIM | Building Information Modeling. Creation and management of digital models describing a building’s physical and functional properties. | Sec. 14.1 |