Intent Recognition in Natural Human-Robot Collaboration

Monday February 17th 2025 at 14:00

Wolfson Building of Mechanical Engineering, Room 206

Abstract:

Human-robot collaboration relies on the ability of robots to intuitively recognize and respond to natural human gestures, which are non-verbal communication methods conveying intent and directives. These gestures, such as pointing or holding, play a crucial role in enabling seamless interaction between humans and robots in shared tasks. Challenges in this domain include variability in environments, differences among users, and the need for robust systems that can operate under dynamic conditions. Addressing these challenges is essential for advancing human-robot collaboration across multiple fields, including healthcare, search and rescue, and industrial automation.

In this research, we proposed innovative frameworks to address key challenges in intent recognition. First, we developed a wearable Force-Myography (FMG) based system for recognizing objects held by users, utilizing the novel Flip-U-Net architecture for robust performance across diverse conditions and multi-user environments. Second, we introduced a framework for robust recognition and estimation of pointing gestures using a single web camera, leveraging a lightweight segmentation-based model to accurately detect gestures and estimate their position and direction. Third, we presented the Ultra-Range Gesture Recognition (URGR) framework, combining a High-Quality Network (HQ-Net) for super-resolution with a Graph-Vision Transformer (GViT) for gesture classification, enabling recognition at distances up to 28 meters. Fourth, we developed the Diffusion in Ultra-Range (DUR) framework to generate high-fidelity synthetic datasets for training gesture recognition models, addressing data scarcity and enhancing performance across diverse scenarios. Finally, we introduced a robust dynamic gesture recognition framework based on the SlowFast-Transformer model, achieving high accuracy in challenging conditions, such as low light and occlusions, further advancing the applicability of gesture recognition systems for real-world applications.

Bio:

Eran Bamani Beeri is a PhD candidate at the School of Mechanical Engineering, Tel Aviv University, under the supervision of Dr. Avishai Sintov. His research focuses on deep learning, computer vision, and human-robot interaction, aiming to develop scalable frameworks for natural and intuitive human-robot collab oration. Eran holds a B.Sc. and M.Sc. in Electronic Engineering, where he specialized in image and signal processing. Eran has extensive experience in research and development in the fields of medical image processing, trajectory estimation, and gesture recognition. His work has been published in leading journals. Eran is expected to graduate in March 2025 and will begin as a post doctoral associate at MIT’s Lab 77, working on rehabilitation robotics within the broader field of human-robot collaboration.

Monday March 5th 2025 at 14:00 

Wolfson Building of Mechanical Engineering, Room 206 

abstarct:

Analogies between quantum and classical systems span numerous domains of physics, from optics and acoustics to condensed matter and fluid dynamics. Surface gravity water waves, in particular, exhibit striking similarities to both quantum mechanics and optics. By studying these analogies, we gain deeper insights into the fundamental behaviors that govern both quantum and classical systems. While quantum mechanics operates at microscopic scales, its principles, such as wave-particle duality, find intriguing classical counterparts, allowing photons and electrons to manifest both wave-like and particle-like properties in ways reminiscent of classical wave dynamics.

This research explores the propagation of surface gravity water waves within a framework inspired by the Schrödinger equation. We successfully observed quantum-like phenomena, including the Kennard cubic phase and Talbot carpets, and extended these studies to nonlinear regimes where fractional Talbot effects disappear due to interference suppression in a nonlinear medium. Additionally, our setup enabled novel observations of Bohmian trajectories, quantum potentials, and Wigner function distributions of wave fields. Furthermore, we investigated the dynamics of the inverted harmonic oscillator, a system central to black hole physics, and metastable states, revealing the dynamics of wave packets in phase space. Our experiments also uncovered logarithmic phase singularities in the water wave field. Lastly, we explored hydrodynamic analogies to the Fermi-Dirac statistics of black holes, where surface gravity wave turbulence exhibited statistical distributions akin to fermionic occupation, providing new perspectives on the thermodynamics and quantum statistical properties of event horizons.

Building upon this foundation, we incorporated the concepts of quantum coherence and decoherence into our study. By emulating decohering quantum evolution through surface gravity water waves, we experimentally replicated Lindblad master equation dynamics, observing the transition from coherent superpositions to classical mixtures. Using cat states as initial conditions, we demonstrated the progressive loss of coherence and reduction of purity over time, mirroring quantum decoherence in open systems. This work establishes a classical platform for investigating decoherence mechanisms that underlie quantum-to-classical transitions, with potential implications for testing theories for quantum technologies.

Beyond surface waves, hydrodynamic analogies extend to particle-wave duality in walking droplet systems, where a bouncing oil droplet interacts with its self-generated wave field. This system exhibits an analogy to the Aharonov-Bohm effect, in which a droplet’s trajectory is influenced by the phase of its surrounding wave field, despite no direct force acting on it—akin to the way an electron’s wavefunction acquires a phase shift in a region with a magnetic vector potential but zero classical field. Such hydrodynamic quantum analogs offer a new lens through which to examine nonlocality and gauge field effects in quantum mechanics.

Our findings open new directions in hydrodynamic analogies for studying quantum scattering, time-dependent potentials, and phase-space representations of black holes.

Short Bio:

Gary completed his B.Sc. in Physics and Astronomy in 2014, as well as his M.Sc. in Experimental Condensed Matter Physics in 2017, at Tel Aviv University. His master’s research was conducted under the supervision of Prof. Tal Schwartz, where he investigated ultrafast chemical physics phenomena and, for the first time, measured the spatial dynamics of exciton-polaritons in organic microcavities.

He earned his Ph.D. in 2023 from Tel Aviv University, where he studied quantum mechanical and optical analogies in surface gravity water waves. His research explored how hydrodynamic systems can emulate quantum effects, bridging classical wave dynamics with quantum optics and condensed matter physics.

Following his Ph.D., he was awarded a prestigious postdoctoral fellowship from the Israeli Ministry of Science, which supported his research at the Center for Quantum Science and Technology. There, he worked on high-dimensional quantum key distribution (QKD) and developed satellite tracking algorithms for secure quantum communication.

He was then invited to MIT as a postdoctoral associate, where he joined the group of Nobel Laureate Wolfgang Ketterle to study ultracold Dysprosium atoms, investigating their applications in quantum many-body physics and quantum simulation. Recently, he was awarded the C.L.E. Moore Instructorship Award and Fellowship at MIT, recognizing his contributions to finding bridges between hydrodynamics and quantum physics.