Author name cluster

Qi Luo

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

11 papers

2 author rows

AAAI Conference 2026 Conference Paper

GUIDE: Gaussian Unified Instance Detection for Enhanced Obstacle Perception in Autonomous Driving

Chunyong Hu
Qi Luo
Jianyun Xu
Song Wang
Qiang Li
Sheng Yang

In the realm of autonomous driving, accurately detecting surrounding obstacles is crucial for effective decision-making. Traditional methods primarily rely on 3D bounding boxes to represent these obstacles, which often fail to capture the complexity of irregularly shaped, real-world objects. To overcome these limitations, we present GUIDE, a novel framework that utilizes 3D Gaussians for instance detection and occupancy prediction. Unlike conventional occupancy prediction methods, GUIDE also offers robust tracking capabilities. Our framework employs a sparse representation strategy, using Gaussian-to-Voxel Splatting to provide fine-grained, instance-level occupancy data without the computational demands associated with dense voxel grids. Experimental validation on the nuScenes dataset demonstrates GUIDE's performance, with an instance occupancy mAP of 21.61, marking a 50% improvement over existing methods, alongside competitive tracking capabilities. GUIDE establishes a new benchmark in autonomous perception systems, effectively combining precision with computational efficiency to better address the complexities of real-world driving environments.

PDF Details DOI

NeurIPS Conference 2025 Conference Paper

Decompile-Bench: Million-Scale Binary-Source Function Pairs for Real-World Binary Decompilation

hanzhuo tan
Xiaolong Tian
Hanrui Qi
Jiaming Liu
Siyi Wang
GAO Zuchen
Qi Luo
Jing Li

Recent advances in LLM-based decompilers have been shown effective to convert low-level binaries into human-readable source code. However, there still lacks a comprehensive benchmark that provides large-scale binary-source function pairs, which is critical for advancing the LLM decompilation technology. Creating accurate binary-source mappings incurs severe issues caused by complex compilation settings and widespread function inlining that obscure the correspondence between binaries and their original source code. Previous efforts have either relied on used contest‐style benchmarks, synthetic binary–source mappings that diverge significantly from the mappings in real world, or partially matched binaries with only code lines or variable names, compromising the effectiveness of analyzing the binary functionality. To alleviate these issues, we introduce Decompile-Bench, the first open-source dataset comprising two million binary-source function pairs condensed from 100 million collected function pairs, i. e. , 450GB of binaries compiled from permissively licensed GitHub projects. For the evaluation purposes, we also developed a benchmark Decompile-Bench-Eval including manually crafted binaries from the well-established HumanEval and MBPP, alongside the compiled GitHub repositories released after 2025 to mitigate data leakage issues. We further explore commonly-used evaluation metrics to provide a thorough assessment of the studied LLM decompilers and find that fine-tuning with Decompile-Bench causes a 20% improvement over previous benchmarks in terms of the re-executability rate. Our code and data has been released in HuggingFace and Github. https: //github. com/anonepo/LLM4Decompile

PDF Details

NeurIPS Conference 2025 Conference Paper

ICLScan: Detecting Backdoors in Black-Box Large Language Models via Targeted In-context Illumination

Xiaoyi Pang
Xuanyi Hao
Song Guo
Qi Luo
Zhibo Wang

The widespread deployment of large language models (LLMs) allows users to access their capabilities via black-box APIs, but backdoor attacks pose serious security risks for API users by hijacking the model behavior. This highlights the importance of backdoor detection technologies to help users audit LLMs before use. However, most existing LLM backdoor defenses require white-box access or costly reverse engineering, limiting their practicality for resource-constrained users. Moreover, they mainly target classification tasks, leaving broader generative scenarios underexplored. To solve the problem, this paper introduces ICLScan, a lightweight framework that exploits targeted in-context learning (ICL) as illumination for backdoor detection in black-box LLMs, which effectively supports generative tasks without additional training or model modifications. ICLScan is based on our finding of backdoor susceptibility amplification: LLMs with pre-embedded backdoors are highly susceptible to new trigger implantation via ICL. Including only a small ratio of backdoor examples (containing ICL-triggered input and target output) in the ICL prompt can induce ICL trigger-specific malicious behavior in backdoored LLMs. ICLScan leverages this phenomenon to detect backdoored LLMs by statistically analyzing whether the success rate of new trigger injection via targeted ICL exceeds a threshold. It requires only multiple queries to estimate the backdoor success rate, overcoming black-box access and computational resource limitations. Extensive experiments across diverse LLMs and backdoor attacks demonstrate ICLScan's effectiveness and efficiency, achieving near-perfect detection performance (precision/recall/F1-score/ROC-AUC all approaching 1) with minimal additional overhead across all settings.

PDF Details

ICRA Conference 2023 Conference Paper

3D Reconstruction of Tibia and Fibula using One General Model and Two X-ray Images

Kai Pan
Shuai Zhang 0029
Liang Zhao 0003
Shoudong Huang
Yanhao Zhang 0003
Hua Wang
Qi Luo

The 3D reconstruction of patient specific bone models plays a crucial role in orthopaedic surgery for clinical evaluation, surgical planning and precise implant design or selection. This paper considers the problem of reconstructing a patient-specific 3D tibia and fibula model from only two 2D X-ray images and one 3D general model segmented from the lower leg CT scans of one randomly selected patient. Currently, the bone 3D reconstruction mainly relies on computed tomography (CT) and magnetic resonance imaging (MRI) scanning-based mode segmentation which result in high radiation exposure or expensive costs. While, the proposed algorithm can accurately and efficiently deform a 3D general model to achieve a patient-specific 3D model that matches the patient's tibia and fibula projections in two 2D X-rays. The algorithm undergoes a preliminary deformation, 2D contour registration, and opti-misation based on the deformation graph that represents the shape deformation of models. Evaluations using simulations, cadaver and in-vivo experiments demonstrate that the proposed algorithm can effectively reconstruct the patient's 3D tibia and fibula surface model with high accuracy.

Details

IROS Conference 2021 Conference Paper

A High-accuracy Framework for Vehicle Dynamic Modeling in Autonomous Driving

Shu Jiang
Yu Wang 0038
Weiman Lin
Yu Cao 0012
Longtao Lin
Jinghao Miao
Qi Luo

Vehicle dynamic models are the key to bridge the gap between simulation and real road test in autonomous driving. An accurate vehicle model allows control algorithms in simulation being transferred to real road test with same quality. In this paper, we present a dynamic model residual correction framework (DRF) for vehicle dynamic modeling. DRF provides a general accuracy improvement framework on existing vehicle dynamic models. On top of any existing open-loop dynamic model, this framework builds a Residual Correction Model (RCM) by integrating deep Neural Networks (NN) with Stochastic Variational Gaussian Process (SVGP) model. RCM takes a sequence of vehicle control commands and dynamic states for a certain time duration as modeling inputs, extracts underlying context from this sequence with deep encoder networks, and predicts open-loop dynamic model prediction errors. Five vehicle dynamic models are derived from DRF via encoder variations. Our contribution is consolidated with evaluation of the absolute trajectory error and the similarity between DRF outputs and the ground truth. Compared to classic rule-based and learning-based vehicle dynamic models, DRF accomplishes as high as 74. 12% to 85. 02% of the absolute trajectory error drop among all DRF variations.

Details

IROS Conference 2020 Conference Paper

FlexiVision: Teleporting the Surgeon's Eyes via Robotic Flexible Endoscope and Head-Mounted Display

Long Qian
Chengzhi Song
Yiwei Jiang
Qi Luo
Xin Ma 0008
Philip Wai Yan Chiu
Zheng Li 0012
Peter Kazanzides

A flexible endoscope introduces more dexterity to the image capturing in endoscopic surgery. However, manual control or automatic control based on instrument tracking does not handle the misorientation between the endoscopic video and the surgeon. We propose an automatic flexible endoscope control method that tracks the surgeon's head with respect to the object in the surgical scene. The robotic flexible endoscope is actuated so that it captures the surgical scene from the same perspective as the surgeon. The surgeon wears a head-mounted display to observe the endoscopic video. The frustum of the flexible endoscope is rendered as an augmented reality overlay to provide surgical guidance. We developed the prototype, FlexiVision, integrating a 6-DOF robotic flexible endoscope based on the da Vinci Research Kit and Microsoft HoloLens. We evaluated the proposed automatic control method via a lesion observation task, and evaluated the AR surgical guidance in a lesion targeting task. The multi-user study results demonstrated that, for both tasks, FlexiVision significantly reduced the completion time (by 59% and 58%), number of errors (by 75% and 95%) and subjective task load level. With FlexiVision, the flexible endoscope could act as the surgeon's eyes teleported into the abdominal cavity of the patient.

Details

IROS Conference 2019 Conference Paper

An Automated Learning-Based Procedure for Large-scale Vehicle Dynamics Modeling on Baidu Apollo Platform

Jiaxuan Xu
Qi Luo
Kecheng Xu
Xiangquan Xiao
Siyang Yu
Jiangtao Hu
Jinghao Miao
Jingao Wang

In the autonomous driving industry, vehicle dynamic models are important to control-in-the-loop simulations. For current commercial self-driving simulators, vehicle dynamic models are expressed explicitly by sophisticated analytical equations, which are accurate but difficult to build and expensive to scale to fleets of vehicles of different brands. In this paper, we introduce a highly automated learning-based vehicle dynamic modeling procedure, which has been deployed on Baidu Apollo self-driving platform, to support cross-vehicle data-driven applications on a large scale. Compared with our previous analytical models, the end-to-end learning-based dynamic models can achieve high accuracy with significantly reduced re-development effort.

Details

IROS Conference 2007 Conference Paper

Modeling and rendering contact torques and twisting effects of deformable objects in haptic Interaction

Qi Luo
Jing Xiao 0001

Contact and deformation modeling for interactive environments has seen many applications, from surgical simulation and training, to virtual prototyping, to teleoperation, etc. , where both visual feedback and haptic feedback are needed in real-time (kHz). In this paper, we consider contacts between a rigid body and an elastic object and address a little studied problem: the modeling and rendering of compliant twisting or rotation of the rigid body on the surface of the elastic object and the associated effects in the deformation of the elastic object. We present a unique strategy to model the contact torques applied to the rigid body and the resulted shape deformation of the elastic object. This strategy extends the general paradigm of contact and deformable modeling introduced by the authors earlier [1] so that not only contact forces but also contact torques, not only compliant translations but also compliant rotations of the rigid body, as well as the resulted deformations of the elastic object can all be simulated in a combined update rate of over lfeHz. The strategy is implemented, and the experimental results confirm its effectiveness and efficiency.

Details

IROS Conference 2006 Conference Paper

Haptic Simulation for Micro/Nano-Scale Optical Fiber Assembly

Qi Luo
Jing Xiao 0001

While there are very high industry demands on optical fiber, little research has been done on the modeling and simulation of the optical fiber assembly. In this paper, the interaction forces in the micro/nano joining step of the optical fiber assembly are modeled. Simulation of assembly in a virtual environment via a haptic device is performed, and experimental results are discussed, which could be used for designing leaning-based controller for automated micro/nano-scale optical fiber assembly

Details

ICRA Conference 2004 Conference Paper

On the Representation of Contact States between Curved Objects

Qi Luo
Ernesto Staffetti
Jing Xiao 0001

Information of high-level, topological contact states is useful and even necessary for a wide range of applications, including many robotic applications. A contact state between two polyhedral objects can be effectively represented as a contact formation in terms of a set of principal contacts between faces, edges, and vertices of the two objects. However, little is done to characterize and represent contact states between curved objects. In order to facilitate the representation of contact states between such objects, we introduce a novel approach to segment the boundary of curved objects based on monotonic changes of curvatures, which we call the curvature monotonic segmentation. We specifically apply this approach to curved 2D and 3D objects with boundary curves or surfaces represented by algebraic polynomials of degrees up to 2. The segmentation yields curvature monotonic faces and edges (or pseudo edges), and vertices (or pseudo vertices). With these faces, (pseudo) edges, and (pseudo) vertices, we effectively extend the concept of contact formation to curved objects to represent high-level, topological contact states between such objects with the same desirable characteristics as the contact formations between polyhedral objects.

Details

ICRA Conference 2003 Conference Paper

Haptic modeling of contact formations and compliant motion

Jing Xiao 0001
Qi Luo
Song You

This paper models the effects of different contact formations and compliant motion on haptic rendering, taking into account friction and gravity. When a held rigid object interacts with another rigid object (in a task such as assembly), the force and moment felt by the operator at any instant depend not only on the contact region but also on the type of the contact state and the type of motion of the held object prior to reaching the current contact configuration, especially in the presence of friction and gravity. We address the modeling of such haptic effects by extending our study for the case of two interacting convex polyhedral rigid bodies to the more general case of interacting non-convex polyhedral objects involving more complex contact formations and compliant motion.

Details