arXiv subjects

Yibo Chen

Publications and source records attributed to Yibo Chen.

36 records · Page 2Linked to original sources

arXiv2025

Cross-Scenario Unified Modeling of User Interests at Billion Scale

User interests on content platforms are inherently diverse, manifesting through complex behavioral patterns across heterogeneous scenarios such as search, feed browsing, and content discovery. Traditional recommendation systems typically prioritize business metric optimization within isolated specific scenarios, neglecting cross-scenario behavioral signals and struggling to integrate advanced techniques like LLMs at billion-scale deployments, which finally limits their ability to capture holistic user interests across platform touchpoints. We propose RED-Rec, an LLM-enhanced hierarchical Recommender Engine for Diversified scenarios, tailored for industry-level content recommendation systems. RED-Rec unifies user interest representations across multiple behavioral contexts by aggregating and synthesizing actions from varied scenarios, resulting in comprehensive item and user modeling. At its core, a two-tower LLM-powered framework enables nuanced, multifaceted representations with deployment efficiency, and a scenario-aware dense mixing and querying policy effectively fuses diverse behavioral signals to capture cross-scenario user intent patterns and express fine-grained, context-specific intents during serving. We validate RED-Rec through online A/B testing on hundreds of millions of users in RedNote through online A/B testing, showing substantial performance gains in both content recommendation and advertisement targeting tasks. We further introduce a million-scale sequential recommendation dataset, RED-MMU, for comprehensive offline training and evaluation. Our work advances unified user modeling, unlocking deeper personalization and fostering more meaningful user engagement in large-scale UGC platforms.

cs.IR↗

arXiv2025

BenCao: An Instruction-Tuned Large Language Model for Traditional Chinese Medicine

Traditional Chinese Medicine (TCM), with a history spanning over two millennia, plays a role in global healthcare. However, applying large language models (LLMs) to TCM remains challenging due to its reliance on holistic reasoning, implicit logic, and multimodal diagnostic cues. Existing TCM-domain LLMs have made progress in text-based understanding but lack multimodal integration, interpretability, and clinical applicability. To address these limitations, we developed BenCao, a ChatGPT-based multimodal assistant for TCM, integrating structured knowledge bases, diagnostic data, and expert feedback refinement. BenCao was trained through natural language instruction tuning rather than parameter retraining, aligning with expert-level reasoning and ethical norms specific to TCM. The system incorporates a comprehensive knowledge base of over 1,000 classical and modern texts, a scenario-based instruction framework for diverse interactions, a chain-of-thought simulation mechanism for interpretable reasoning, and a feedback refinement process involving licensed TCM practitioners. BenCao connects to external APIs for tongue-image classification and multimodal database retrieval, enabling dynamic access to diagnostic resources. In evaluations across single-choice question benchmarks and multimodal classification tasks, BenCao achieved superior accuracy to general-domain and TCM-domain models, particularly in diagnostics, herb recognition, and constitution classification. The model was deployed as an interactive application on the OpenAI GPTs Store, accessed by nearly 1,000 users globally as of October 2025. This study demonstrates the feasibility of developing a TCM-domain LLM through natural language-based instruction tuning and multimodal integration, offering a practical framework for aligning generative AI with traditional medical reasoning and a scalable pathway for real-world deployment.

cs.CL↗

arXiv2025

CQ-DINO: Mitigating Gradient Dilution via Category Queries for Vast Vocabulary Object Detection

With the exponential growth of data, traditional object detection methods are increasingly struggling to handle vast vocabulary object detection tasks effectively. We analyze two key limitations of classification-based detectors: positive gradient dilution, where rare positive categories receive insufficient learning signals, and hard negative gradient dilution, where discriminative gradients are overwhelmed by numerous easy negatives. To address these challenges, we propose CQ-DINO, a category query-based object detection framework that reformulates classification as a contrastive task between object queries and learnable category queries. Our method introduces image-guided query selection, which reduces the negative space by adaptively retrieving top-K relevant categories per image via cross-attention, thereby rebalancing gradient distributions and facilitating implicit hard example mining. Furthermore, CQ-DINO flexibly integrates explicit hierarchical category relationships in structured datasets (e.g., V3Det) or learns implicit category correlations via self-attention in generic datasets (e.g., COCO). Experiments demonstrate that CQ-DINO achieves superior performance on the challenging V3Det benchmark (surpassing previous methods by 2.1% AP) while maintaining competitiveness in COCO. Our work provides a scalable solution for real-world detection systems requiring wide category coverage. The code is publicly at https://github.com/FireRedTeam/CQ-DINO.

cs.CV↗

arXiv2025

Toward Super Agent System with Hybrid AI Routers

AI Agents powered by Large Language Models are transforming the world through enormous applications. A super agent has the potential to fulfill diverse user needs, such as summarization, coding, and research, by accurately understanding user intent and leveraging the appropriate tools to solve tasks. However, to make such an agent viable for real-world deployment and accessible at scale, significant optimizations are required to ensure high efficiency and low cost. This position paper presents a design of the Super Agent System powered by the hybrid AI routers. Upon receiving a user prompt, the system first detects the intent of the user, then routes the request to specialized task agents with the necessary tools or automatically generates agentic workflows. In practice, most applications directly serve as AI assistants on edge devices such as phones and robots. As different language models vary in capability and cloud-based models often entail high computational costs, latency, and privacy concerns, we then explore the hybrid mode where the router dynamically selects between local and cloud models based on task complexity. Finally, we introduce the blueprint of an on-device super agent enhanced with cloud. With advances in multi-modality models and edge hardware, we envision that most computations can be handled locally, with cloud collaboration only as needed. Such architecture paves the way for super agents to be seamlessly integrated into everyday life in the near future.

cs.AI↗

arXiv2025

Through Silicon Via Aware Design Planning for Thermally Efficient 3-D Integrated Circuits

3-D integrated circuits (3-D ICs) offer performance advantages due to their increased bandwidth and reduced wire-length enabled by through-silicon-via structures (TSVs). Traditionally TSVs have been considered to improve the thermal conductivity in the vertical direction. However, the lateral thermal blockage effect becomes increasingly important for TSV via farms (a cluster of TSV vias used for signal bus connections between layers) because the TSV size and pitch continue to scale in μm range and the metal to insulator ratio becomes smaller. Consequently, dense TSV farms can create lateral thermal blockages in thinned silicon substrate and exacerbate the local hotspots. In this paper, we propose a thermal-aware via farm placement technique for 3-D ICs to minimize lateral heat blockages caused by dense signal bus TSV structures.

cs.AR↗

arXiv2025

Arrested on heating: controlling the motility of active droplets by temperature

Self-propelling active matter relies on the conversion of energy from the undirected, nanoscopic scale to directed, macroscopic motion. One of the challenges in the design of synthetic active matter lies in the control of dynamic states, or motility gaits. Here, we present an experimental system of self-propelling droplets with thermally controllable and reversible dynamic states, from unsteady over meandering to persistent to arrested motion. These states are known to depend on the Péclet number of the molecular process powering the motion, which we can now tune by using a temperature sensitive mixture of surfactants as propulsion fuel. We quantify the droplet dynamics by analyzing flow and chemical fields for the individual states, comparing them to canonical models for autophoretic particles. In the context of these models, we experimentally observe, in situ, the fundamental first broken symmetry that translates an isotropic, immotile base state to self-propelled motility.

physics.flu-dyn↗

arXiv2024

BeadSight: An Inexpensive Tactile Sensor Using Hydro-Gel Beads

In robotic manipulation, tactile sensors are indispensable, especially when dealing with soft objects, objects of varying dimensions, or those out of the robot's direct line of sight. Traditional tactile sensors often grapple with challenges related to cost and durability. To address these issues, our study introduces a novel approach to visuo-tactile sensing with an emphasis on economy and replacablity. Our proposed sensor, BeadSight, uses hydro-gel beads encased in a vinyl bag as an economical, easily replaceable sensing medium. When the sensor makes contact with a surface, the deformation of the hydrogel beads is observed using a rear camera. This observation is then passed through a U-net Neural Network to predict the forces acting on the surface of the bead bag, in the form of a pressure map. Our results show that the sensor can accurately predict these pressure maps, detecting the location and magnitude of forces applied to the surface. These abilities make BeadSight an effective, inexpensive, and easily replaceable tactile sensor, ideal for many robotics applications.

cs.RO↗

arXiv2023

Buoyancy-driven attraction of active droplets

Active oil droplets in a liquid are believed to repel due to the Marangoni effect, while buoyancy effects caused by the density difference between the droplets, diffusing product, and ambient fluid are usually overlooked. Recent experiments have observed active droplet clustering phenomena due to buoyancy-driven convection (Kruger et al. Eur. Phys. J. E, vol. 39, 2016, pp.1-9). In this study, we numerically analyze the buoyancy effect in addition to Marangoni flow, characterized by Peclet number $Pe$. The buoyancy effects originate from (i) the density difference between the droplet and the ambient liquid, which is characterized by Galileo number $Ga$, and (ii) the density difference between the diffusing product (i.e. filled micelles) and the ambient liquid, characterized by a solutal Rayleigh number $Ra$. We analyze how the attracting and repulsing behavior depends on the control parameters $Pe$, $Ga$, and $Ra$. We find that while Marangoni flow causes repulsion, the buoyancy effect leads to attraction, and even collisions can take place at high Ra. We also observe a delayed collision as $Ga$ increases. Moreover, we derive that the attracting velocity, characterized by a Reynolds number $Re_d$, is proportional to $Ra^{1/4}/(l/R)$, where $l/R$ is the normalized distance by radius between neighboring droplets. Finally, we obtain repulsive velocity, characterized by $Re_{rep}$, as proportional to $PeRa^{-0.38}$. The balance of attractive and repulsive effects results in $Pe \sim Ra^{0.63}$, which agrees with the transition curve between regimes with and without collision.

physics.flu-dyn↗

arXiv2021

GraphPAS: Parallel Architecture Search for Graph Neural Networks

Graph neural architecture search has received a lot of attention as Graph Neural Networks (GNNs) has been successfully applied on the non-Euclidean data recently. However, exploring all possible GNNs architectures in the huge search space is too time-consuming or impossible for big graph data. In this paper, we propose a parallel graph architecture search (GraphPAS) framework for graph neural networks. In GraphPAS, we explore the search space in parallel by designing a sharing-based evolution learning, which can improve the search efficiency without losing the accuracy. Additionally, architecture information entropy is adopted dynamically for mutation selection probability, which can reduce space exploration. The experimental result shows that GraphPAS outperforms state-of-art models with efficiency and accuracy simultaneously.

cs.LG↗

arXiv2021

Droplet plume emission during plasmonic bubble growth in ternary liquids

Plasmonic bubbles are of great relevance in numerous applications, including catalytic reactions, micro/nanomanipulation of molecules or particles dispersed in liquids, and cancer therapeutics. So far, studies have been focused on bubble nucleation in pure liquids. Here we investigate plasmonic bubble nucleation in ternary liquids consisting of ethanol, water, and trans-anethole oil which can show the so-called ouzo-effect. We find that oil (trans-anethole) droplet plumes are produced around the growing plasmonic bubbles. The nucleation of the microdroplets and their organization in droplet plumes is due to the symmetry breaking of the ethanol concentration field during the selective evaporation of ethanol from the surrounding ternary liquids into the growing plasmonic bubbles. Numerical simulations show the existence of a critical Marangoni number Ma (the ratio between solutal advection rate and the diffusion rate), above which the symmetry breaking of the ethanol concentration field occurs, leading to the emission of the droplet plumes. The numerical results agree with the experimental observation that more plumes are emitted with increasing ethanol-water relative weight ratios and hence Ma. Our findings on the droplet plume formation reveal the rich phenomena of plasmonic bubble nucleation in multicomponent liquids and help to pave the way to achieve enhanced mixing in multicomponent liquids in chemical, pharmaceutical, and cosmetic industries.

physics.flu-dyn↗

arXiv2021

Collective entrainment and confinement amplify transport by schooling micro-swimmers

Micro-swimmers can serve as cargo carriers that move deep inside complex flow networks. When a school collectively entrains the surrounding fluid, their transport capacity can be enhanced. This effect is quantified with good agreement between experiments with self-propelled droplets and a confined Brinkman squirmer model. The volume of liquid entrained can be much larger than the droplet itself, amplifying the effective cargo capacity over an order of magnitude, even for dilute schools. Hence, biological and engineered swimmers can efficiently transport materials into confined environments.

cond-mat.soft↗

arXiv2021

Instabilities driven by diffusio-phoretic flow on catalytic surfaces

We theoretically and numerically investigate the instabilities driven by diffusiophoretic flow, caused by a solutal concentration gradient along a reacting surface. The important control parameter is the Peclet number Pe, which quantifies the ratio of the solutal advection rate to the diffusion rate. First, we study the diffusiophoretic flow on a catalytic plane in two dimensions. From a linear stability analysis, we obtain that for Pe larger than 8pi, mass transport by convection overtakes that by diffusion, and a symmetry-breaking mode arises, which is consistent with numerical results. For even larger Pe, non-linear terms become important. For Pe larger than 16pi, multiple concentration plumes are emitted from the catalytic plane, which eventually merges into a single larger one. When Pe is even larger, there are continuous emissions and merging events of the concentration plumes. This newly-found flow state reflects the non-linear saturation of the system. The critical Peclet number for the transition to this state depends on Schmidt number Sc. In the second part of the paper, we conduct three-dimensional simulations for spherical catalytic particles, and beyond a critical Peclet number again find continuous plume emission and plume merging, now leading to chaotic motion of the phoretic particle. Our results thus help to understand the experimentally observed chaotic motion of catalytic particles in the high Pe regime.

physics.flu-dyn↗

arXiv2020

Self-propelling Microdroplets Generated and Sustained by Liquid-liquid Phase Separation in Confined Spaces

Flow transport in confined spaces is ubiquitous in technological processes, ranging from separation and purification of pharmaceutical ingredients by microporous membranes and drug delivery in biomedical treatment to chemical and biomass conversion in catalyst-packed reactors and carbon dioxide sequestration. In this work, we suggest a distinct pathway for enhanced liquid transport in a confined space via self-propelling microdroplets. These microdroplets can form spontaneously from localized liquid-liquid phase separation as a ternary mixture is diluted by a diffusing poor solvent. High speed images reveal how the microdroplets grow, break up and propel rapidly along the solid surface, with a maximal velocity up to ~160 um/s, in response to a sharp concentration gradient resulting from phase separation. The microdroplet self-propulsion induces a replenishing flow between the walls of the confined space towards the location of phase separation, which in turn drives the mixture out of equilibrium and leads to a repeating cascade of events. Our findings on the complex and rich phenomena of self-propelling droplets suggest an effective approach to enhanced flow motion of multicomponent liquid mixtures within confined spaces for time effective separation and smart transport processes.

cond-mat.soft↗

arXiv2020

Dissolution of microdroplets in a sparsely miscible liquid confined by leaky walls

When a water droplet is deposited within a sparsely miscible liquid medium such as certain oils, the droplet surprisingly vanishes, even in a confined geometry. Such a phenomenon has crucial consequences for multiphase flows in which confined nano- and/or picoliter droplets are considered. We report here experiments of microdroplet dissolution in microchannels that reveal an enhancement of the shrinkage of confined water microdroplets in oil due to the permeability of the walls - made of polydimethilsiloxane (PDMS) - and a delay when collective effects are present. The system is first modelled assuming that the dissolution of the droplets in its surrounding liquid follows the Epstein-Plesset solution of the diffusion equation. The dissolution of small isolated droplets can indeed be described by this solution of the diffusion equation, while the vanishing of droplets larger than a certain critical value and those closer to other droplets requires numerical simulations which take into account the boundary conditions of the confined system, the neighbouring droplets and interestingly, the evaporative water vapour flux through the PDMS. Our results thus reveal the important role of the water solubility in oil and most remarkably, of the water vapour transport through permeable walls.

physics.flu-dyn↗

arXiv2019

Zoom-In-to-Check: Boosting Video Interpolation via Instance-level Discrimination

We propose a light-weight video frame interpolation algorithm. Our key innovation is an instance-level supervision that allows information to be learned from the high-resolution version of similar objects. Our experiment shows that the proposed method can generate state-of-the-art results across different datasets, with fractional computation resources (time and memory) of competing methods. Given two image frames, a cascade network creates an intermediate frame with 1) a flow-warping module that computes coarse bi-directional optical flow and creates an interpolated image via flow-based warping, followed by 2) an image synthesis module to make fine-scale corrections. In the learning stage, object detection proposals are generated on the interpolated image.Lower resolution objects are zoomed into, and the learning algorithms using an adversarial loss trained on high-resolution objects to guide the system towards the instance-level refinement corrects details of object shape and boundaries.

cs.CV↗

arXiv2019

The bilateral solver for quality estimation based multi-focus image fusion

In this work, a fast Bilateral Solver for Quality Estimation Based multi-focus Image Fusion method (BS-QEBIF) is proposed. The all-in-focus image is generated by pixel-wise summing up the multi-focus source images with their focus-levels maps as weights. Since the visual quality of an image patch is highly correlated with its focus level, the focus-level maps are preliminarily obtained based on visual quality scores, as pre-estimations. However, the pre-estimations are not ideal. Thus the fast bilateral solver is then adopted to smooth the pre-estimations, and edges in the multi-focus source images can be preserved simultaneously. The edge-preserving smoothed results are utilized as final focus-level maps. Moreover, this work provides a confidence-map solution for the unstable fusion in the focus-level-changed boundary regions. Experiments were conducted on $25$ pairs of source images. The proposed BS-QEBIF outperforms the other $13$ fusion methods objectively and subjectively. The all-in-focus image produced by the proposed method can well maintain the details in the multi-focus source images and does not suffer from any residual errors. Experimental results show that BS-QEBIF can handle the focus-level-changed boundary regions without any blocking, ringing and blurring artifacts.

cs.MM↗

arXiv2018

Realtime Time Synchronized Event-based Stereo

In this work, we propose a novel event based stereo method which addresses the problem of motion blur for a moving event camera. Our method uses the velocity of the camera and a range of disparities to synchronize the positions of the events, as if they were captured at a single point in time. We represent these events using a pair of novel time synchronized event disparity volumes, which we show remove motion blur for pixels at the correct disparity in the volume, while further blurring pixels at the wrong disparity. We then apply a novel matching cost over these time synchronized event disparity volumes, which both rewards similarity between the volumes while penalizing blurriness. We show that our method outperforms more expensive, smoothing based event stereo methods, by evaluating on the Multi Vehicle Stereo Event Camera dataset.

cs.CV↗

arXiv2016

DoraPicker: An Autonomous Picking System for General Objects

Robots that autonomously manipulate objects within warehouses have the potential to shorten the package delivery time and improve the efficiency of the e-commerce industry. In this paper, we present a robotic system that is capable of both picking and placing general objects in warehouse scenarios. Given a target object, the robot autonomously detects it from a shelf or a table and estimates its full 6D pose. With this pose information, the robot picks the object using its gripper, and then places it into a container or at a specified location. We describe our pick-and-place system in detail while highlighting our design principles for the warehouse settings, including the perception method that leverages knowledge about its workspace, three grippers designed to handle a large variety of different objects in terms of shape, weight and material, and grasp planning in cluttered scenarios. We also present extensive experiments to evaluate the performance of our picking system and demonstrate that the robot is competent to accomplish various tasks in warehouse settings, such as picking a target item from a tight space, grasping different objects from the shelf, and performing pick-and-place tasks on the table.

cs.RO↗