1st edition · ECCV 2024

Archived edition

Eyes of the Future

Integrating Computer Vision in Smart Eyewear (ICVSE)

The first workshop in the series established a dedicated forum for smart-eyewear computer vision, connecting algorithmic advances with computation, energy, privacy and usability constraints.

12accepted papers

70+attendees

5invited speakers

September 29, 2024 · Morning

Program

Invited talks were interleaved with speed poster presentations and a one-hour poster and networking session.

09:00–09:10

Opening by the organizers

Welcome

09:10–09:30

Jakob Engel · Meta Reality Labs

Invited talk

09:30–09:50

Giovanni Maria Farinella · University of Catania

Invited talk

09:50–10:30

Speed poster presentations

Contributed papers

10:30–11:30

Coffee break and poster session

Posters & networking

11:30–11:50

Elisabetta Farella · Fondazione Bruno Kessler

Invited talk

11:50–12:10

Yoichi Sato · The University of Tokyo

Invited talk

12:10–12:30

Enkelejda Kasneci · Technical University of Munich

Invited talk

12:30–13:00

Awards ceremony and closing remarks

Closing

Invited talks

Invited Speakers

The program covered spatial AI, egocentric vision, hardware-aware tinyML, visual attention and gaze guidance in extended reality.

Jakob Engel

Meta Reality Labs

TalkSpatial AI for Contextual AI

Giovanni Maria Farinella

University of Catania

TalkProcedure Understanding from Egocentric Videos

Elisabetta Farella

Fondazione Bruno Kessler

TalkHardware-Aware Scaling in tinyML

Yoichi Sato

The University of Tokyo

TalkUnderstanding Egocentric Visual Attention and Actions

Enkelejda Kasneci

Technical University of Munich

TalkImperceptible Gaze Guidance in Extended Reality

Recognition

Awards

Best Paper Awards

Three submissions recognized

The workshop recognized three submissions based on novelty, impact and research quality during the closing session.

Workshop contributions

Accepted Papers

Twelve papers spanning eye tracking, event-based sensing, visual-inertial SLAM, efficient transformers, pose estimation and egocentric action recognition.

Helios: An extremely low power event-based gesture recognition for always-on smart eyewearPrarthana Bhattacharyya; Joshua Mitton; Ryan F. Page; Richard G. Vigars; Louis Berridge; Owen Morgan; Kemi Jacobs; Ben Menzies; Gabriel Homewood; Christopher J. Mair; Taru Muhonen; Rory Clark; Paolo Baesso; Dave Trickett; Iain Wallace
Towards Resource-aware Visual Inertial SLAMGiovanni Affatato; Marco Paracchini; Francesca Palermo; Diana Trojaniello; Tommaso Ongarello; Marco Marcon; Stefano Tubaro
Ultra-Efficient On-Device Object Detection on AI-Integrated Smart Glasses With TinyissimoYOLOJulian Moosmann; Pietro Bonazzi; Philipp Mayer; Yawei Li; Sizhen Bian; Luca Benini; Michele Magno
Evaluating Human Pose Estimation Algorithms for Resource-Constrained Smart Eyewear DeviceHao Quan; Francesca Palermo; Simone Mentasti; Diana Trojaniello; Matteo Matteucci
Selective Merging Schedule for Vision Transformers on Edge DevicesShivam Aggarwal; Pranav Venkatram; Tulika Mitra
Towards Real-Time Online Egocentric Action Recognition on Smart EyewearRiccardo Santambrogio; Federico Caspani; Greta Corti; Francesca Palermo; Simone Mentasti; Diana Trojaniello; Matteo Matteucci
Normalized Validity Scores for DNNs in Regression based Eye Feature Extraction and Real-Time Models for the Raspberry PiWolfgang Fuhl
Real-Time 2nd-order Gaze MetricsAndrew Duchowski; Krzysztof Krejtz; Izabela Krejtz
CondSeg: Ellipse Estimation of Pupil and Iris via Conditioned SegmentationZhuang Jia; Jiangfan Deng; Liying Chi; Xiang Long; Daniel Kang Du
Towards Unsupervised Eye-Region Segmentation for Eye TrackingJiangfan Deng; Zhuang Jia; Zhaoxue Wang; Xiang Long; Daniel Kang Du
Towards Low-power, High-frequency Gaze Direction Tracking with an Event-cameraYvonne Vullers; Luna Gava; Arren Glover; Chiara Bartolozzi
High-frequency near-eye ground truth for event-based eye trackingAndrea Simpsi; Andrea Aspesi; Simone Mentasti; Luca Merigo; Tommaso Ongarello; Matteo Matteucci