arXiv subjects

Duo Lu

Publications and source records attributed to Duo Lu.

15 recordsLinked to original sources

arXiv2026

Credo: Reusable Declarative Primitives for Agentic Workflows

An LLM application depends on both a model and a harness: the program that determines what each call sees, how many calls to make, and which answers to trust. Coding agents can now discover strong harnesses by searching over candidate programs, but the resulting artifact is an opaque block of imperative code whose logical steps, runtime signals, physical execution decisions, and prompt strategies remain implicit and task-specific, forcing subsequent tasks to start the harness search process from scratch. The potential for reuse, however, is substantial. A searched harness encodes significant knowledge, such as the logical steps that work, the signals that matter, the physical operator decisions that adapt execution, and the prompt strategies that are effective, yet this knowledge is buried in imperative code with no inspectable or reusable structure, nor does it carry any provenance or metadata. Credo addresses this problem by recovering a structured declarative description of a searched harness, tagging each extracted primitive with relevant metadata, and cataloguing all of it with provenance. A compiler can then bind stored primitives to generate harnesses for new tasks without having to start the search over from scratch. This paper provides preliminary results demonstrating the potential of our approach and lays out a related research agenda that the database community is well-positioned to tackle, including cost-based compilation over declarative catalogs and catalog maintenance under model and workload drift.

cs.AI↗

arXiv2026

Declarative Data Services: Structured Agentic Discovery for Composing Data Systems

Agentic discovery has shown that LLM-driven search can find novel algorithms, designs, and code under benchmark conditions. Translating the paradigm to multi-system data backends surfaces a harder problem: the search space is heterogeneous, the verifier is whether a deployed stack actually runs, and composition knowledge is unevenly captured in pretraining. Unbounded agentic discovery, a coding agent iterating on failure-log feedback, fails to converge consistently on a working stack even when iteration and explicit composition knowledge are added. We propose Declarative Data Services (DDS), an architecture for structured agentic discovery of data-system compositions from declarative user intent. The framework owns four typed contracts at successive layers (intent, operator DAG, per-system skills, runtime attribution) that decompose the global search into bounded sub-searches; sub-agents search each typed space, while the framework provides the channels by which knowledge flows forward as inline skill citations and errors route backward as typed signals. As a proof of life on a trading-backend workload, DDS converges where unbounded discovery does not; runtime failures become skill patches that the next deployment cites inline. We position this as an early prototype reporting lessons from real-world data-system composition.

cs.AI↗

arXiv2026

Credo: Declarative Control of LLM Pipelines via Beliefs and Policies

Agentic AI systems are becoming commonplace in domains that require long-lived, stateful decision-making in continuously evolving conditions. As such, correctness depends not only on the output of individual model calls, but also on how to best adapt when incorporating new evidence or revising prior conclusions. However, existing frameworks rely on imperative control loops, ephemeral memory, and prompt-embedded logic, making agent behavior opaque, brittle, and difficult to verify. This paper introduces Credo, which represents semantic state as beliefs and regulates behavior using declarative policies defined over these beliefs. This design supports adaptive, auditable, and composable execution through a database-backed semantic control plane. We showcase these concepts in a decision-control scenario, where beliefs and policies declaratively guide critical execution choices (e.g., model selection, retrieval, corrective re-execution), enabling dynamic behavior without requiring any changes to the underlying pipeline code.

cs.AI↗

arXiv2026

An In-Depth Study of Filter-Agnostic Vector Search on a PostgreSQL Database System: [Experiments and Analysis]

Filtered Vector Search (FVS) is critical for supporting semantic search and GenAI applications in modern database systems. However, existing research most often evaluates algorithms in specialized libraries, making optimistic assumptions that do not align with enterprise-grade database systems. Our work challenges this premise by demonstrating that in a production-grade database system, commonly made assumptions do not hold, leading to performance characteristics and algorithmic trade-offs that are fundamentally different from those observed in isolated library settings. This paper presents the first in-depth analysis of filter-agnostic FVS algorithms within a production PostgreSQL-compatible system. We systematically evaluate post-filtering and inline-filtering strategies across a wide range of selectivities and correlations. Our central finding is that the optimal algorithm is not dictated by the cost of distance computations alone, but that system-level overheads that come from both distance computations and filter operations (like page accesses and data retrieval) play a significant role. We demonstrate that graph-based approaches (such as NaviX/ACORN) can incur prohibitive numbers of filter checks and system-level overheads, compared with clustering-based indexes such as ScaNN, often canceling out their theoretical benefits in real-world database environments. Ultimately, our findings provide the database community with crucial insights and practical guidelines, demonstrating that the optimal choice for a filter-agnostic FVS algorithm is not absolute, but rather a system-aware decision contingent on the interplay between workload characteristics and the underlying costs of data access in a real-world database architecture.

cs.DB↗

arXiv2026

Domain Expansion: A Latent Space Construction Framework for Multi-Task Learning

Training a single network with multiple objectives often leads to conflicting gradients that degrade shared representations, forcing them into a compromised state that is suboptimal for any single task--a problem we term latent representation collapse. We introduce Domain Expansion, a framework that prevents these conflicts by restructuring the latent space itself. Our framework uses a novel orthogonal pooling mechanism to construct a latent space where each objective is assigned to a mutually orthogonal subspace. We validate our approach across diverse benchmarks--including ShapeNet, MPIIGaze, and Rotated MNIST--on challenging multi-objective problems combining classification with pose and gaze estimation. Our experiments demonstrate that this structure not only prevents collapse but also yields an explicit, interpretable, and compositional latent space where concepts can be directly manipulated.

cs.LG↗

arXiv2025

Making Prompts First-Class Citizens for Adaptive LLM Pipelines

Modern LLM pipelines increasingly resemble complex data-centric applications: they retrieve data, correct errors, call external tools, and coordinate interactions between agents. Yet, the central element controlling this entire process -- the prompt -- remains a brittle, opaque string that is entirely disconnected from the surrounding program logic. This disconnect fundamentally limits opportunities for reuse, optimization, and runtime adaptivity. In this paper, we describe our vision and an initial design of SPEAR (Structured Prompt Execution and Adaptive Refinement), a new approach to prompt management that treats prompts as first-class citizens in the execution model. Specifically, SPEAR enables: (1) structured prompt management, with prompts organized into versioned views to support introspection and reasoning about provenance; (2) adaptive prompt refinement, whereby prompts can evolve dynamically during execution based on runtime feedback; and (3) policy-driven control, a mechanism for the specification of automatic prompt refinement logic as when-then rules. By tackling the problem of runtime prompt refinement, SPEAR plays a complementary role in the vast ecosystem of existing prompt optimization frameworks and semantic query processing engines. We describe a number of related optimization opportunities unlocked by the SPEAR model, and our preliminary results demonstrate the strong potential of this approach.

cs.DB↗

arXiv2025

MC-BEVRO: Multi-Camera Bird Eye View Road Occupancy Detection for Traffic Monitoring

Single camera 3D perception for traffic monitoring faces significant challenges due to occlusion and limited field of view. Moreover, fusing information from multiple cameras at the image feature level is difficult because of different view angles. Further, the necessity for practical implementation and compatibility with existing traffic infrastructure compounds these challenges. To address these issues, this paper introduces a novel Bird's-Eye-View road occupancy detection framework that leverages multiple roadside cameras to overcome the aforementioned limitations. To facilitate the framework's development and evaluation, a synthetic dataset featuring diverse scenes and varying camera configurations is generated using the CARLA simulator. A late fusion and three early fusion methods were implemented within the proposed framework, with performance further enhanced by integrating backgrounds. Extensive evaluations were conducted to analyze the impact of multi-camera inputs and varying BEV occupancy map sizes on model performance. Additionally, a real-world data collection pipeline was developed to assess the model's ability to generalize to real-world environments. The sim-to-real capabilities of the model were evaluated using zero-shot and few-shot fine-tuning, demonstrating its potential for practical application. This research aims to advance perception systems in traffic monitoring, contributing to improved traffic management, operational efficiency, and road safety.

cs.CV↗

arXiv2024

Roundabout Dilemma Zone Data Mining and Forecasting with Trajectory Prediction and Graph Neural Networks

Traffic roundabouts, as complex and critical road scenarios, pose significant safety challenges for autonomous vehicles. In particular, the encounter of a vehicle with a dilemma zone (DZ) at a roundabout intersection is a pivotal concern. This paper presents an automated system that leverages trajectory forecasting to predict DZ events, specifically at traffic roundabouts. Our system aims to enhance safety standards in both autonomous and manual transportation. The core of our approach is a modular, graph-structured recurrent model that forecasts the trajectories of diverse agents, taking into account agent dynamics and integrating heterogeneous data, such as semantic maps. This model, based on graph neural networks, aids in predicting DZ events and enhances traffic management decision-making. We evaluated our system using a real-world dataset of traffic roundabout intersections. Our experimental results demonstrate that our dilemma forecasting system achieves a high precision with a low false positive rate of 0.1. This research represents an advancement in roundabout DZ data mining and forecasting, contributing to the assurance of intersection safety in the era of autonomous vehicles.

cs.CV↗

arXiv2024

SynTraC: A Synthetic Dataset for Traffic Signal Control from Traffic Monitoring Cameras

This paper introduces SynTraC, the first public image-based traffic signal control dataset, aimed at bridging the gap between simulated environments and real-world traffic management challenges. Unlike traditional datasets for traffic signal control which aim to provide simplified feature vectors like vehicle counts from traffic simulators, SynTraC provides real-style images from the CARLA simulator with annotated features, along with traffic signal states. This image-based dataset comes with diverse real-world scenarios, including varying weather and times of day. Additionally, SynTraC also provides different reward values for advanced traffic signal control algorithms like reinforcement learning. Experiments with SynTraC demonstrate that it is still an open challenge to image-based traffic signal control methods compared with feature-based control methods, indicating our dataset can further guide the development of future algorithms. The code for this paper can be found in \url{https://github.com/DaRL-LibSignal/SynTraC}.SynTraC

cs.AI↗

arXiv2023

SKoPe3D: A Synthetic Dataset for Vehicle Keypoint Perception in 3D from Traffic Monitoring Cameras

Intelligent transportation systems (ITS) have revolutionized modern road infrastructure, providing essential functionalities such as traffic monitoring, road safety assessment, congestion reduction, and law enforcement. Effective vehicle detection and accurate vehicle pose estimation are crucial for ITS, particularly using monocular cameras installed on the road infrastructure. One fundamental challenge in vision-based vehicle monitoring is keypoint detection, which involves identifying and localizing specific points on vehicles (such as headlights, wheels, taillights, etc.). However, this task is complicated by vehicle model and shape variations, occlusion, weather, and lighting conditions. Furthermore, existing traffic perception datasets for keypoint detection predominantly focus on frontal views from ego vehicle-mounted sensors, limiting their usability in traffic monitoring. To address these issues, we propose SKoPe3D, a unique synthetic vehicle keypoint dataset generated using the CARLA simulator from a roadside perspective. This comprehensive dataset includes generated images with bounding boxes, tracking IDs, and 33 keypoints for each vehicle. Spanning over 25k images across 28 scenes, SKoPe3D contains over 150k vehicle instances and 4.9 million keypoints. To demonstrate its utility, we trained a keypoint R-CNN model on our dataset as a baseline and conducted a thorough evaluation. Our experiments highlight the dataset's applicability and the potential for knowledge transfer between synthetic and real-world data. By leveraging the SKoPe3D dataset, researchers and practitioners can overcome the limitations of existing datasets, enabling advancements in vehicle keypoint detection for ITS.

cs.CV↗

arXiv2023

CAROM -- Vehicle Localization and Traffic Scene Reconstruction from Monocular Cameras on Road Infrastructures

Traffic monitoring cameras are powerful tools for traffic management and essential components of intelligent road infrastructure systems. In this paper, we present a vehicle localization and traffic scene reconstruction framework using these cameras, dubbed as CAROM, i.e., "CARs On the Map". CAROM processes traffic monitoring videos and converts them to anonymous data structures of vehicle type, 3D shape, position, and velocity for traffic scene reconstruction and replay. Through collaborating with a local department of transportation in the United States, we constructed a benchmarking dataset containing GPS data, roadside camera videos, and drone videos to validate the vehicle tracking results. On average, the localization error is approximately 0.8 m and 1.7 m within the range of 50 m and 120 m from the cameras, respectively.

cs.RO↗

arXiv2023

CAROM Air -- Vehicle Localization and Traffic Scene Reconstruction from Aerial Videos

Road traffic scene reconstruction from videos has been desirable by road safety regulators, city planners, researchers, and autonomous driving technology developers. However, it is expensive and unnecessary to cover every mile of the road with cameras mounted on the road infrastructure. This paper presents a method that can process aerial videos to vehicle trajectory data so that a traffic scene can be automatically reconstructed and accurately re-simulated using computers. On average, the vehicle localization error is about 0.1 m to 0.3 m using a consumer-grade drone flying at 120 meters. This project also compiles a dataset of 50 reconstructed road traffic scenes from about 100 hours of aerial videos to enable various downstream traffic analysis applications and facilitate further road traffic related research. The dataset is available at https://github.com/duolu/CAROM.

cs.CV↗

arXiv2022

Evaluating Persistent Memory Range Indexes: Part Two [Extended Version]

Scalable persistent memory (PM) has opened up new opportunities for building indexes that operate and persist data directly on the memory bus, potentially enabling instant recovery, low latency and high throughput. When real PM hardware (Intel Optane DCPMM) first became available, previous work evaluated PM indexes proposed in the pre-Optane era. Since then, newer indexes based on real PM have appeared, but it is unclear how they compare to each other and to previous proposals, and what further challenges remain. This paper addresses these issues by analyzing and experimentally evaluating state-of-the-art PM range indexes built for real PM. We find newer designs inherited past techniques with new improvements, but they do not necessarily outperform pre-Optane era proposals. Moreover, PM indexes are often also very competitive or even outperform indexes tailored for DRAM, highlighting the potential of using a unified design for both PM and DRAM. Functionalitywise, these indexes still lack good support for variable-length keys and handling NUMA effect. Based on our findings, we distill new design principles and highlight future directions.

cs.DB↗

arXiv2019

FMHash: Deep Hashing of In-Air-Handwriting for User Identification

Many mobile systems and wearable devices, such as Virtual Reality (VR) or Augmented Reality (AR) headsets, lack a keyboard or touchscreen to type an ID and password for signing into a virtual website. However, they are usually equipped with gesture capture interfaces to allow the user to interact with the system directly with hand gestures. Although gesture-based authentication has been well-studied, less attention is paid to the gesture-based user identification problem, which is essentially an input method of account ID and an efficient searching and indexing method of a database of gesture signals. In this paper, we propose FMHash (i.e., Finger Motion Hash), a user identification framework that can generate a compact binary hash code from a piece of in-air-handwriting of an ID string. This hash code enables indexing and fast search of a large account database using the in-air-handwriting by a hash table. To demonstrate the effectiveness of the framework, we implemented a prototype and achieved >99.5% precision and >92.6% recall with exact hash code match on a dataset of 200 accounts collected by us. The ability of hashing in-air-handwriting pattern to binary code can be used to achieve convenient sign-in and sign-up with in-air-handwriting gesture ID on future mobile and wearable systems connected to the Internet.

cs.CV↗

arXiv2018

FMCode: A 3D In-the-Air Finger Motion Based User Login Framework for Gesture Interface

Applications using gesture-based human-computer interface require a new user login method with gestures because it does not have a traditional input method to type a password. However, due to various challenges, existing gesture-based authentication systems are generally considered too weak to be useful in practice. In this paper, we propose a unified user login framework using 3D in-air-handwriting, called FMCode. We define new types of features critical to distinguish legitimate users from attackers and utilize Support Vector Machine (SVM) for user authentication. The features and data-driven models are specially designed to accommodate minor behavior variations that existing gesture authentication methods neglect. In addition, we use deep neural network approaches to efficiently identify the user based on his or her in-air-handwriting, which avoids expansive account database search methods employed by existing work. On a dataset collected by us with over 100 users, our prototype system achieves 0.1% and 0.5% best Equal Error Rate (EER) for user authentication, as well as 96.7% and 94.3% accuracy for user identification, using two types of gesture input devices. Compared to existing behavioral biometric systems using gesture and in-air-handwriting, our framework achieves the state-of-the-art performance. In addition, our experimental results show that FMCode is capable to defend against client-side spoofing attacks, and it performs persistently in the long run. These results and discoveries pave the way to practical usage of gesture-based user login over the gesture interface.

cs.CV↗