arXiv subjects

Kaifeng Zhao

Publications and source records attributed to Kaifeng Zhao.

18 recordsLinked to original sources

arXiv2026

Enhanced Atomic Magnetometry with a Pinched Spin State

The sensitivity of an atomic spin sensor is fundamentally constrained by the quantum Fisher information (QFI) of the probe state. The pinched state, $|{F,0}\rangle$, has an $(F+1)$-fold larger single-particle QFI for transverse-field sensing than the stretched state, $|{F,F}\rangle$, corresponding to a predicted $\sqrt{F+1}$ reduction in the spin-projection-noise-limited field uncertainty. We show that this advantage can be accessed in an rf sensing scheme by engineering a spectrum symmetric about $m=0$, yielding an ideal $(F+1)$-fold enhancement of magnetic-field response. For $^{87}\mathrm{Rb}$ with $F=2$, we observe up to 2.7-fold response enhancement. Implementing this approach in a compact $0.8~\mathrm{cm}^3$ anti-relaxation-coated multipass cell, we realize a single-beam, dual-axis, zero-field magnetometer, achieving $13$--$23$~$\mathrm{fT}/\sqrt{\mathrm{Hz}}$ over $3$--$100$~Hz at room temperature.

quant-ph↗

arXiv2026

ADAPT: Agile Diffusion Action Priors for Robust and Steerable Online Text-Driven Humanoid Control

We present ADAPT, an end-to-end framework for interactive, text-conditioned humanoid whole-body control. Unlike dominant text-to-motion pipelines that generate kinematic motions for a separate tracker, ADAPT solves language control with an end-to-end closed-loop control framework, where the robot must continuously respond to changing commands while maintaining balance, natural motion, and smooth transitions. ADAPT learns a diffusion-based action prior from text-labeled humanoid state-action trajectories, enabling diverse motion skills to be directly executed from language commands. To improve long-horizon robustness and smooth prompt switching, we train a lightweight residual reinforcement learning policy on top of the frozen diffusion controller. We further show that the same diffusion policy can be reused as a steerable text-conditioned motion prior for downstream task adaptation. Experiments demonstrate robust language-grounded skill execution, smooth interactive transitions, and style-preserving downstream control.

cs.RO↗

arXiv2026

ARDY: Autoregressive Diffusion with Hybrid Representation for Interactive Human Motion Generation

Generating realistic 3D human motions in real-time within interactive applications is key for animation, simulation, and humanoid robotics. While recent offline motion generation approaches offer precise control via text and kinematic constraints, they lack the inference speed required for interactive settings. Conversely, existing online methods enable real-time synthesis but often sacrifice controllability or struggle with complex text semantics and long-horizon goals due to limited context windows. In this work, we introduce ARDY, a streaming generation framework that bridges this gap by enabling high-fidelity motion generation controllable via online text prompts and flexible kinematic constraints. ARDY employs a hybrid representation that combines explicit root features with a latent body embedding, balancing precise trajectory control with efficient generative learning. We propose a two-stage autoregressive transformer denoiser that features variable history context and supports conditioning on flexible, long-horizon kinematic constraints. By training on a large-scale motion capture dataset and being directly conditioned on text labels and kinematic constraints sampled from ground truth poses, ARDY natively learns controllable generation that supports online prompting and flexible long-horizon goals. Extensive evaluations on the HumanML3D benchmark and the large-scale, high-fidelity Bones Rigplay dataset demonstrate ARDY's high motion quality and constraint adherence, validating the efficacy of our key architectural decisions. Finally, we demonstrate the method's practical versatility through an interactive demo featuring dynamic text control, diverse keyframe pose constraints, path following, and interactive locomotion control via mouse and keyboard. Supplementary video results, code, and model releases can be found at https://research.nvidia.com/labs/sil/projects/ardy/.

cs.GR↗

arXiv2026

FID Magnetometer Based on Paraffin-Coated Planar Reflective Multipass Cells

We demonstrate a paraffin-coated planar reflective multipass vapor cell for compact optical atomic magnetometry. The cell has an internal volume of $12 \times 12 \times 8~\mathrm{mm}^3$ and supports 20 optical passes with a total transmittance exceeding $65\%$, while maintaining a longitudinal spin-relaxation time of $^{87}\mathrm{Rb}$ longer than $1~\mathrm{s}$. The planar geometry provides spatially separated input and output beams, enabling compact optical integration. A single-cell free-induction-decay (FID) magnetometer reaches $10~\mathrm{pT}/\sqrt{\mathrm{Hz}}$ in the geomagnetic-field range, presently limited by current-source noise in the field coils. A two-cell differential configuration achieves a sensitivity of $\sim 28~\mathrm{fT}/\sqrt{\mathrm{Hz}}$ over the $1$--$15~\mathrm{Hz}$ band for bias fields of $0.3$--$0.7~\mathrm{G}$. These results establish that paraffin-coated planar multipass cells offer high optical depth, long coherence times, and an integration-friendly platform for ultrasensitive magnetometry.

physics.optics↗

arXiv2026

MotionBricks: Scalable Real-Time Motions with Modular Latent Generative Model and Smart Primitives

Despite transformative advances in generative motion synthesis, real-time interactive motion control remains dominated by traditional techniques. In this work, we identify two key challenges in bridging research and production: 1) Real-time scalability: Industry applications demand real-time generation of a vast repertoire of motion skills, while generative methods exhibit significant degradation in quality and scalability under real-time computation constraints, and 2) Integration: Industry applications demand fine-grained multi-modal control involving velocity commands, style selection, and precise keyframes, a need largely unmet by existing text- or tag-driven models. To overcome these limitations, we introduce MotionBricks: a large-scale, real-time generative framework with a two-fold solution. First, we propose a large-scale modular latent generative backbone tailored for robust real-time motion generation, effectively modeling a dataset of over 350,000 motion clips with a single model. Second, we introduce smart primitives that provide a unified, robust, and intuitive interface for authoring both navigation and object interaction. Applications can be designed in a plug-and-play manner like assembling bricks without expert animation knowledge. Quantitatively, we show that MotionBricks produces state-of-the-art motion quality on open-source and proprietary datasets of various scales, while also achieving a real-time throughput of 15,000 FPS with 2ms latency. We demonstrate the flexibility and robustness of MotionBricks in a complete production-level animation demo, covering navigation and object-scene interaction across various styles with a unified model. To showcase our framework's application beyond animation, we deploy MotionBricks on the Unitree G1 humanoid robot to demonstrate its flexibility and generalization for real-time robotic control.

cs.RO↗

arXiv2025

EgoM2P: Egocentric Multimodal Multitask Pretraining

Understanding multimodal signals in egocentric vision, such as RGB video, depth, camera poses, and gaze, is essential for applications in augmented reality, robotics, and human-computer interaction, enabling systems to better interpret the camera wearer's actions, intentions, and surrounding environment. However, building large-scale egocentric multimodal and multitask models presents unique challenges. Egocentric data are inherently heterogeneous, with large variations in modality coverage across devices and settings. Generating pseudo-labels for missing modalities, such as gaze or head-mounted camera trajectories, is often infeasible, making standard supervised learning approaches difficult to scale. Furthermore, dynamic camera motion and the complex temporal and spatial structure of first-person video pose additional challenges for the direct application of existing multimodal foundation models. To address these challenges, we introduce a set of efficient temporal tokenizers and propose EgoM2P, a masked modeling framework that learns from temporally-aware multimodal tokens to train a large, general-purpose model for egocentric 4D understanding. This unified design supports multitasking across diverse egocentric perception and synthesis tasks, including gaze prediction, egocentric camera tracking, and monocular depth estimation from egocentric video, and also serves as a generative model for conditional egocentric video synthesis. Across these tasks, EgoM2P matches or outperforms specialist models while being an order of magnitude faster. We will fully open-source EgoM2P to support the community and advance egocentric vision research. Project page: https://egom2p.github.io/.

cs.CV↗

arXiv2025

VolumetricSMPL: A Neural Volumetric Body Model for Efficient Interactions, Contacts, and Collisions

Parametric human body models play a crucial role in computer graphics and vision, enabling applications ranging from human motion analysis to understanding human-environment interactions. Traditionally, these models use surface meshes, which pose challenges in efficiently handling interactions with other geometric entities, such as objects and scenes, typically represented as meshes or point clouds. To address this limitation, recent research has explored volumetric neural implicit body models. However, existing works are either insufficiently robust for complex human articulations or impose high computational and memory costs, limiting their widespread use. To this end, we introduce VolumetricSMPL, a neural volumetric body model that leverages Neural Blend Weights (NBW) to generate compact, yet efficient MLP decoders. Unlike prior approaches that rely on large MLPs, NBW dynamically blends a small set of learned weight matrices using predicted shape- and pose-dependent coefficients, significantly improving computational efficiency while preserving expressiveness. VolumetricSMPL outperforms prior volumetric occupancy model COAP with 10x faster inference, 6x lower GPU memory usage, enhanced accuracy, and a Signed Distance Function (SDF) for efficient and differentiable contact modeling. We demonstrate VolumetricSMPL's strengths across four challenging tasks: (1) reconstructing human-object interactions from in-the-wild images, (2) recovering human meshes in 3D scenes from egocentric views, (3) scene-constrained motion synthesis, and (4) resolving self-intersections. Our results highlight its broad applicability and significant performance and efficiency gains.

cs.CV↗

arXiv2025

DartControl: A Diffusion-Based Autoregressive Motion Model for Real-Time Text-Driven Motion Control

Text-conditioned human motion generation, which allows for user interaction through natural language, has become increasingly popular. Existing methods typically generate short, isolated motions based on a single input sentence. However, human motions are continuous and can extend over long periods, carrying rich semantics. Creating long, complex motions that precisely respond to streams of text descriptions, particularly in an online and real-time setting, remains a significant challenge. Furthermore, incorporating spatial constraints into text-conditioned motion generation presents additional challenges, as it requires aligning the motion semantics specified by text descriptions with geometric information, such as goal locations and 3D scene geometry. To address these limitations, we propose DartControl, in short DART, a Diffusion-based Autoregressive motion primitive model for Real-time Text-driven motion control. Our model effectively learns a compact motion primitive space jointly conditioned on motion history and text inputs using latent diffusion models. By autoregressively generating motion primitives based on the preceding history and current text input, DART enables real-time, sequential motion generation driven by natural language descriptions. Additionally, the learned motion primitive space allows for precise spatial motion control, which we formulate either as a latent noise optimization problem or as a Markov decision process addressed through reinforcement learning. We present effective algorithms for both approaches, demonstrating our model's versatility and superior performance in various motion synthesis tasks. Experiments show our method outperforms existing baselines in motion realism, efficiency, and controllability. Video results are available on the project page: https://zkf1997.github.io/DART/.

cs.CV↗

arXiv2024

EgoGen: An Egocentric Synthetic Data Generator

Understanding the world in first-person view is fundamental in Augmented Reality (AR). This immersive perspective brings dramatic visual changes and unique challenges compared to third-person views. Synthetic data has empowered third-person-view vision models, but its application to embodied egocentric perception tasks remains largely unexplored. A critical challenge lies in simulating natural human movements and behaviors that effectively steer the embodied cameras to capture a faithful egocentric representation of the 3D world. To address this challenge, we introduce EgoGen, a new synthetic data generator that can produce accurate and rich ground-truth training data for egocentric perception tasks. At the heart of EgoGen is a novel human motion synthesis model that directly leverages egocentric visual inputs of a virtual human to sense the 3D environment. Combined with collision-avoiding motion primitives and a two-stage reinforcement learning approach, our motion synthesis model offers a closed-loop solution where the embodied perception and movement of the virtual human are seamlessly coupled. Compared to previous works, our model eliminates the need for a pre-defined global path, and is directly applicable to dynamic environments. Combined with our easy-to-use and scalable data generation pipeline, we demonstrate EgoGen's efficacy in three tasks: mapping and localization for head-mounted cameras, egocentric camera tracking, and human mesh recovery from egocentric views. EgoGen will be fully open-sourced, offering a practical solution for creating realistic egocentric training data and aiming to serve as a useful tool for egocentric computer vision research. Refer to our project page: https://ego-gen.github.io/.

cs.CV↗

arXiv2023

Scalable Probabilistic Forecasting in Retail with Gradient Boosted Trees: A Practitioner's Approach

The recent M5 competition has advanced the state-of-the-art in retail forecasting. However, we notice important differences between the competition challenge and the challenges we face in a large e-commerce company. The datasets in our scenario are larger (hundreds of thousands of time series), and e-commerce can afford to have a larger assortment than brick-and-mortar retailers, leading to more intermittent data. To scale to larger dataset sizes with feasible computational effort, firstly, we investigate a two-layer hierarchy and propose a top-down approach to forecasting at an aggregated level with less amount of series and intermittency, and then disaggregating to obtain the decision-level forecasts. Probabilistic forecasts are generated under distributional assumptions. Secondly, direct training at the lower level with subsamples can also be an alternative way of scaling. Performance of modelling with subsets is evaluated with the main dataset. Apart from a proprietary dataset, the proposed scalable methods are evaluated using the Favorita dataset and the M5 dataset. We are able to show the differences in characteristics of the e-commerce and brick-and-mortar retail datasets. Notably, our top-down forecasting framework enters the top 50 of the original M5 competition, even with models trained at a higher level under a much simpler setting.

cs.LG↗

arXiv2023

Synthesizing Diverse Human Motions in 3D Indoor Scenes

We present a novel method for populating 3D indoor scenes with virtual humans that can navigate in the environment and interact with objects in a realistic manner. Existing approaches rely on training sequences that contain captured human motions and the 3D scenes they interact with. However, such interaction data are costly, difficult to capture, and can hardly cover all plausible human-scene interactions in complex environments. To address these challenges, we propose a reinforcement learning-based approach that enables virtual humans to navigate in 3D scenes and interact with objects realistically and autonomously, driven by learned motion control policies. The motion control policies employ latent motion action spaces, which correspond to realistic motion primitives and are learned from large-scale motion capture data using a powerful generative motion model. For navigation in a 3D environment, we propose a scene-aware policy with novel state and reward designs for collision avoidance. Combined with navigation mesh-based path-finding algorithms to generate intermediate waypoints, our approach enables the synthesis of diverse human motions navigating in 3D indoor scenes and avoiding obstacles. To generate fine-grained human-object interactions, we carefully curate interaction goal guidance using a marker-based body representation and leverage features based on the signed distance field (SDF) to encode human-scene proximity relations. Our method can synthesize realistic and diverse human-object interactions (e.g.,~sitting on a chair and then getting up) even for out-of-distribution test scenarios with different object shapes, orientations, starting body positions, and poses. Experimental results demonstrate that our approach outperforms state-of-the-art methods in terms of both motion naturalness and diversity. Code and video results are available at: https://zkf1997.github.io/DIMOS.

cs.CV↗

arXiv2022

Compositional Human-Scene Interaction Synthesis with Semantic Control

Synthesizing natural interactions between virtual humans and their 3D environments is critical for numerous applications, such as computer games and AR/VR experiences. Our goal is to synthesize humans interacting with a given 3D scene controlled by high-level semantic specifications as pairs of action categories and object instances, e.g., "sit on the chair". The key challenge of incorporating interaction semantics into the generation framework is to learn a joint representation that effectively captures heterogeneous information, including human body articulation, 3D object geometry, and the intent of the interaction. To address this challenge, we design a novel transformer-based generative model, in which the articulated 3D human body surface points and 3D objects are jointly encoded in a unified latent space, and the semantics of the interaction between the human and objects are embedded via positional encoding. Furthermore, inspired by the compositional nature of interactions that humans can simultaneously interact with multiple objects, we define interaction semantics as the composition of varying numbers of atomic action-object pairs. Our proposed generative model can naturally incorporate varying numbers of atomic interactions, which enables synthesizing compositional human-scene interactions without requiring composite interaction data. We extend the PROX dataset with interaction semantic labels and scene instance segmentation to evaluate our method and demonstrate that our method can generate realistic human-scene interactions with semantic control. Our perceptual study shows that our synthesized virtual humans can naturally interact with 3D scenes, considerably outperforming existing methods. We name our method COINS, for COmpositional INteraction Synthesis with Semantic Control. Code and data are available at https://github.com/zkf1997/COINS.

cs.CV↗

arXiv2022

Intense, wideband optical waveform generation by self-balanced amplification of fiber electro-optical sideband modulation

We demonstrate a simple method to obtain accurate optical waveforms with a GHz-level programmable modulation bandwidth and Watt-level output power for wideband optical control of free atoms and molecules. Arbitrary amplitude and phase modulations are transferred from microwave to light with a low-power fiber electro-optical modulator. The sub-milliWatt optical sideband is co-amplified with the optical carrier in a power-balanced fashion through a tapered semiconductor amplifier (TSA). By automatically keeping TSA near saturation in a quasi-continuous manner, typical noise channels associated with pulsed high-gain amplifications are efficiently suppressed. As an example application, we demonstrate interleaved cooling and trapping of two rubidium isotopes with coherent nanosecond pulses.

physics.optics↗

arXiv2021

Zero-field spin noise spectrum of an alkali vapor with strong spin-exchange coupling

We study the zero-field optical spin noise spectra (OSN) for a thermal state hot 87Rb vapor with strong spin exchange coupling. Our main findings are: (1) The OSN spectrum consists of two components representing a positive and a negative hyperfine spin correlation (HSC), the relative power of which varies dramatically with the detuning frequency of the probe. (2) There exist two polar frequencies at which the OSN spectrum is completely polarized with one type of HSC. (3) At the limit of far detuning, the power ratio of the positive and negative HSC component of the OSN is 4:5. (4) The total power of the OSN is independent of the strength of SE coupling. (5) We give a simple way of deriving the OSN using the eigensolution of the density matrix equation.

quant-ph↗

arXiv2019

Observing zero-field spin dynamics with spin noise in a pi-pulse modulated field

Spin noise spectroscopic study of spin dynamics in a zero magnetic field is commonly frustrated by the dominating 1/f noise. We show that in a pi-pulse modulated magnetic field, spin noise spectrum centered one-half of the field modulation frequency reveals spin dynamics in zero-fields, free of any low-frequency noises.

physics.atom-ph↗

arXiv2018

Shapley Value Methods for Attribution Modeling in Online Advertising

This paper re-examines the Shapley value methods for attribution analysis in the area of online advertising. As a credit allocation solution in cooperative game theory, Shapley value method directly quantifies the contribution of online advertising inputs to the advertising key performance indicator (KPI) across multiple channels. We simplify its calculation by developing an alternative mathematical formulation. The new formula significantly improves the computational efficiency and therefore extends the scope of applicability. Based on the simplified formula, we further develop the ordered Shapley value method. The proposed method is able to take into account the order of channels visited by users. We claim that it provides a more comprehensive insight by evaluating the attribution of channels at different stages of user conversion journeys. The proposed approaches are illustrated using a real-world online advertising campaign dataset.

econ.EM↗

arXiv2017

Revenue-based Attribution Modeling for Online Advertising

This paper examines and proposes several attribution modeling methods that quantify how revenue should be attributed to online advertising inputs. We adopt and further develop relative importance method, which is based on regression models that have been extensively studied and utilized to investigate the relationship between advertising efforts and market reaction (revenue). Relative importance method aims at decomposing and allocating marginal contributions to the coefficient of determination (R^2) of regression models as attribution values. In particular, we adopt two alternative submethods to perform this decomposition: dominance analysis and relative weight analysis. Moreover, we demonstrate an extension of the decomposition methods from standard linear model to additive model. We claim that our new approaches are more flexible and accurate in modeling the underlying relationship and calculating the attribution values. We use simulation examples to demonstrate the superior performance of our new approaches over traditional methods. We further illustrate the value of our proposed approaches using a real advertising campaign dataset.

econ.EM↗

arXiv2013

Bayesian Quantile Regression for Partially Linear Additive Models

In this article, we develop a semiparametric Bayesian estimation and model selection approach for partially linear additive models in conditional quantile regression. The asymmetric Laplace distribution provides a mechanism for Bayesian inferences of quantile regression models based on the check loss. The advantage of this new method is that nonlinear, linear and zero function components can be separated automatically and simultaneously during model fitting without the need of pre-specification or parameter tuning. This is achieved by spike-and-slab priors using two sets of indicator variables. For posterior inferences, we design an effective partially collapsed Gibbs sampler. Simulation studies are used to illustrate our algorithm. The proposed approach is further illustrated by applications to two real data sets.

stat.CO↗