Searcharxiv⌕ Search

arXiv subjects

Man Cao

Publications and source records attributed to Man Cao.

2 recordsLinked to original sources

ROSE: A Shared Cross-Embodiment Facial-Motion Latent Space for Humanoid Robot Expression

Natural, coherent full-face dynamic expressions are central to making humanoid robots socially approachable, yet existing systems commonly realize only limited lip motion or scripted expressions. The limitation is both physical and algorithmic. A compact head leaves little room for densely packed, independently controllable actuation; meanwhile, learned systems often generate an avatar video, landmarks, or facial parameters and then infer actuator commands through visual retargeting. This render-and-invert path adds latency and artifacts, and it does not reliably transfer generated behavior to a physical robot. We present ROSE (RObot facial expression through Shared motion Embeddings), a framework that integrates biomimetic hardware, cross-embodiment motion representation learning, and multimodal motion generation. On the hardware side, we build a flexible tendon-sheath transmission structure that integrates 34 actuators into a 19.5 cm biomimetic head. By decoupling actuator placement from facial action points, it enables coordinated full-face expression within a compact volume. On the algorithmic side, ROSE-VAE learns an identity-independent shared facial-motion latent space from human, digital-human, and robot observations. ROSE-Flow directly maps text, audio, or their combination to trajectories in this motion space without using video as a generator input. ROSE-ACT then converts these trajectories into constrained actuator commands. Our system demonstrates how high-density biomimetic hardware and a shared executable facial-motion representation can jointly unify cross-embodiment learning, multimodal generation, and physical robot expression.

cs.RO↗

Continuous and discrete Painleve equations arising from the gap probability distribution of the finite $n$ Gaussian Unitary Ensembles

In this paper we study the gap probability problem in the Gaussian Unitary Ensembles of $n$ by $n$ matrices : The probability that the interval $J := (-a,a)$ is free of eigenvalues. In the works of Tracy and Widom, Adler and Van Moerbeke and Forrester and Witte on this subject, it has been shown that two Painleve type differential equations arise in this context. The first is the Jimbo-Miwa-Okomoto $σ-$form and the second is a particular Painleve IV. Using the ladder operator technique of orthogonal polynomials we derive three quantities associated with the gap probability, denoted by $σ_n(a)$, $R_n(a)$ and $r_n(a)$, and show that each one satisfying a second order, non-linear, differential equation as well as a second order, non-linear difference equation. In particular, in addition to providing an elementary derivation of the aforementioned $σ-$form and Painleve IV we show that the quantity $r_n(a)$ satisfies a particular case of Chazy's second degree second order differential equation. For the discrete equations we show that the quantity $r_n(a)$ satisfies a particular form of the modified discrete Painleve II equation obtained by Grammaticos and Ramani in the context of Backlund transformations. We also derive second order second degree difference equations for the quantities $R_n(a)$ and $σ_n(a)$.

math.CA↗