Sensing primer
Compare signal characteristics, hardware cost, deployment constraints, and privacy trade-offs across RF, acoustic, thermal, ToF, and IMU sensing.
ACM UbiComp / ISWC 2026 · Half-day Tutorial
Towards Open-Sourced Datasets, Benchmarks, and Platforms for Multi-Modal Sensing
A structured and hands-on introduction to open datasets, benchmark design, multi-modal learning, and practical resources for ubiquitous sensing.
Format
3-hour tutorial
Date
To be announced
Venue
ACM UbiComp / ISWC 2026
Overview
Why this tutorial?
Ubiquitous computing is moving from bespoke, single-sensor systems toward robust multi-modal perception. Newly released datasets and benchmarks are making it possible to compare sensing modalities, reproduce results, and develop more generalizable models.
This tutorial combines concise lectures, live sensor demonstrations, interactive checkpoints, and a guided Python/PyTorch notebook session. Participants will learn how to compare datasets fairly, select suitable evaluation protocols, and run unimodal and multi-modal baselines through a common workflow.
What you will learn
From sensing signals to reproducible benchmarks
Dataset landscape
Navigate MM-Fi, OctoNet, XRF55, mRI, OPERAnet, and other representative resources using a common comparison framework.
Learning paradigms
Understand early, intermediate, late, and attention-based fusion, cross-modal supervision, missing-modality robustness, and sensor foundation models.
Hands-on practice
Run and interpret prepared baselines, enable fusion, change or mask a modality, and evaluate performance in a guided notebook.
Representative program
Three hours of lectures, interaction, and practice
The exact start time will follow the official UbiComp / ISWC 2026 schedule.
Opening
Motivation, scope, learning goals, and tutorial format.
Module I · Sensing modalities primer
RF, acoustic, thermal, ToF, and inertial sensing, followed by a short signal-inspection checkpoint and live sensor demonstrations.
Module II · Datasets and benchmarks
Cross-dataset comparison covering modalities, tasks, scale, labels, synchronization, licensing, and evaluation protocols, followed by a dataset-selection exercise.
Coffee break and demo browsing
Informal Q&A and hands-on exploration of representative sensors.
Module III · Models and learning paradigms
Per-modality encoders, fusion strategies, cross-modal learning, missing-modality robustness, and a model-comparison checkpoint.
Module IV · Guided hands-on benchmark
Load prepared samples, run a baseline, enable fusion, mask or swap a modality, and interpret benchmark metrics using a dataset-agnostic notebook.
Closing panel
Open research questions, community wishlist, and Q&A.
Open tutorial materials
Designed for reuse after the conference
Hands-on notebook
A self-contained Jupyter/Colab notebook with prepared samples, pretrained checkpoints, modality adapters, fusion, and evaluation.
Dataset comparison toolkit
A cross-dataset comparison matrix, dataset-selection checklist, licensing references, and common evaluation template.
Slides and reading list
Public slides and a curated reading list covering datasets, benchmark design, fusion, cross-modal learning, and sensing models.
Materials status:
Tutorial resources are under preparation and will be released before the conference.
Who should attend?
Open to newcomers and active researchers
The tutorial is intended for graduate students, postdoctoral researchers, faculty members, and industry researchers or engineers working in ubiquitous computing, mobile and wearable sensing, HCI, signal processing, and applied machine learning.
Prerequisites
- Working knowledge of Python and a framework such as PyTorch
- Basic familiarity with signal processing and machine learning
- No prior experience with a specific sensing modality is required
What to bring
A laptop with Python 3.10+, PyTorch, and approximately 5 GB of free disk space. A fully runnable Google Colab version will also be provided.
Organizers
The University of Hong Kong
Xie Zhang
Spatial sensing intelligence and privacy-preserving multi-modal sensing.
Weiying Hou
Wireless sensing systems and low-altitude technologies.
Chenshu Wu
AIoT, wireless sensing, mobile systems, and ubiquitous computing.