Yiting Chen

Qitian Wu, Yiting Chen, Chenxiao Yang et al.

Learning on graphs, where instance nodes are inter-connected, has become one of the central problems for deep learning, as relational structures are pervasive and induce data inter-dependence which hinders trivial adaptation of existing approaches that assume inputs to be i.i.d.~sampled. However, current models mostly focus on improving testing performance of in-distribution data and largely ignore the potential risk w.r.t. out-of-distribution (OOD) testing samples that may cause negative outcome if the prediction is overconfident on them. In this paper, we investigate the under-explored problem, OOD detection on graph-structured data, and identify a provably effective OOD discriminator based on an energy function directly extracted from graph neural networks trained with standard classification loss. This paves a way for a simple, powerful and efficient OOD detection model for GNN-based learning on graphs, which we call GNNSafe. It also has nice theoretical properties that guarantee an overall distinguishable margin between the detection scores for in-distribution and OOD samples, which, more critically, can be further strengthened by a learning-free energy belief propagation scheme. For comprehensive evaluation, we introduce new benchmark settings that evaluate the model for detecting OOD data from both synthetic and real distribution shifts (cross-domain graph shifts and temporal graph shifts). The results show that GNNSafe achieves up to $17.0\%$ AUROC improvement over state-of-the-arts and it could serve as simple yet strong baselines in such an under-developed area.

Junjia Liu, Yiting Chen, Zhipeng Dong et al.

This letter describes an approach to achieve well-known Chinese cooking art stir-fry on a bimanual robot system. Stir-fry requires a sequence of highly dynamic coordinated movements, which is usually difficult to learn for a chef, let alone transfer to robots. In this letter, we define a canonical stir-fry movement, and then propose a decoupled framework for learning this deformable object manipulation from human demonstration. First, the dual arms of the robot are decoupled into different roles (a leader and follower) and learned with classical and neural network-based methods separately, then the bimanual task is transformed into a coordination problem. To obtain general bimanual coordination, we secondly propose a Graph and Transformer based model -- Structured-Transformer, to capture the spatio-temporal relationship between dual-arm movements. Finally, by adding visual feedback of content deformation, our framework can adjust the movements automatically to achieve the desired stir-fry effect. We verify the framework by a simulator and deploy it on a real bimanual Panda robot system. The experimental results validate our framework can realize the bimanual robot stir-fry motion and have the potential to extend to other deformable objects with bimanual coordination.

Pu Feng, Junkang Liang, Size Wang et al.

In multi-agent reinforcement learning (MARL), the Centralized Training with Decentralized Execution (CTDE) framework is pivotal but struggles due to a gap: global state guidance in training versus reliance on local observations in execution, lacking global signals. Inspired by human societal consensus mechanisms, we introduce the Hierarchical Consensus-based Multi-Agent Reinforcement Learning (HC-MARL) framework to address this limitation. HC-MARL employs contrastive learning to foster a global consensus among agents, enabling cooperative behavior without direct communication. This approach enables agents to form a global consensus from local observations, using it as an additional piece of information to guide collaborative actions during execution. To cater to the dynamic requirements of various tasks, consensus is divided into multiple layers, encompassing both short-term and long-term considerations. Short-term observations prompt the creation of an immediate, low-layer consensus, while long-term observations contribute to the formation of a strategic, high-layer consensus. This process is further refined through an adaptive attention mechanism that dynamically adjusts the influence of each consensus layer. This mechanism optimizes the balance between immediate reactions and strategic planning, tailoring it to the specific demands of the task at hand. Extensive experiments and real-world applications in multi-robot systems showcase our framework's superior performance, marking significant advancements over baselines.

Jiacheng Xu, Heting Gao, Liufei Xie et al.

Human speech conveys expressiveness beyond linguistic content, including personality, mood, or performance elements, such as a comforting tone or humming a song, which we formalize as role-playing and singing. We present VITA-QinYu, the first expressive end-to-end (E2E) spoken language model (SLM) that goes beyond natural conversation to support both role-playing and singing generation. VITA-QinYu adopts a hybrid speech-text paradigm that extends interleaved text-audio modeling with multi-codebook audio tokens, a design enabling richer paralinguistic representation while preserving a clear separation between modalities to avoid interference. We further develop a comprehensive data generation pipeline to synthesize a total of 15.8K hours of natural conversation, role-playing, and singing data for training. VITA-QinYu demonstrates superior expressiveness, outperforming peer SLMs by 7 percentage points on objective role-playing benchmarks, and surpassing peer models by 0.13 points on a 5-point MOS scale for singing. Simultaneously, it achieves state-of-the-art conversational accuracy and fluency, exceeding prior SLMs by 1.38 and 4.98 percentage points on the C3 and URO benchmarks, respectively. We open-source our code and models and provide an easy-to-use demo with full-stack support for streaming and full-duplex interaction.

Yiting Chen, Zhanpeng Zhou, Junchi Yan

The behavior of neural networks still remains opaque, and a recently widely noted phenomenon is that networks often achieve similar performance when initialized with different random parameters. This phenomenon has attracted significant attention in measuring the similarity between features learned by distinct networks. However, feature similarity could be vague in describing the same feature since equivalent features hardly exist. In this paper, we expand the concept of equivalent feature and provide the definition of what we call functionally equivalent features. These features produce equivalent output under certain transformations. Using this definition, we aim to derive a more intrinsic metric for the so-called feature complexity regarding the redundancy of features learned by a neural network at each layer. We offer a formal interpretation of our approach through the lens of category theory, a well-developed area in mathematics. To quantify the feature complexity, we further propose an efficient algorithm named Iterative Feature Merging. Our experimental results validate our ideas and theories from various perspectives. We empirically demonstrate that the functionally equivalence widely exists among different features learned by the same neural network and we could reduce the number of parameters of the network without affecting the performance.The IFM shows great potential as a data-agnostic model prune method. We have also drawn several interesting empirical findings regarding the defined feature complexity.

Yiting Chen, Pol Mestres, Emiliano Dall'Anese et al.

Control barrier function (CBF)-based safety filters provide a systematic way to enforce state constraints, but they can significantly alter the closed-loop dynamics induced by a nominal, stabilizing controller. In particular, the resulting safety-filtered system may exhibit undesirable behaviors including limit cycles, unbounded trajectories, and undesired equilibria. This paper develops a policy optimization framework to maximally enhance the stability properties of safety-filtered controllers. Focusing on linear systems with linear nominal controllers, we jointly parameterize the nominal feedback gain and safety-filter components, and optimize them using trajectory-based objectives computed from closed-loop rollouts. To ensure that the nominal controller remains stabilizing throughout training, we encode Lyapunov-based stability conditions as smooth scalar constraints and enforce them using robust safe gradient flow. This guarantees feasibility of the stability constraints along the optimization iterates and therefore avoids instability during training. Numerical experiments on obstacle-avoidance problems show that the proposed approach can remove asymptotically stable undesired equilibria and improve convergence behavior while maintaining forward invariance of the safe set.

Tianyi Xu, Yiting Chen, Henger Li et al.

We formalize sequential decision-making with information acquisition as the probing-augmented user-centric selection (PUCS) framework, where a learner first probes a subset of arms to obtain side information on resources and rewards, and then assigns $K$ plays to $M$ arms. PUCS covers applications such as ridesharing, wireless scheduling, and content recommendation, in which both resources and payoffs are initially unknown and probing is costly. For the offline setting with known distributions, we present a greedy probing algorithm with a constant-factor approximation guarantee $ζ= (e-1)/(2e-1)$. For the online setting with unknown distributions, we introduce OLPA, a stochastic combinatorial bandit algorithm that achieves a regret bound $\mathcal{O}(\sqrt{T} + \ln^{2} T)$. We also prove a lower bound $Ω(\sqrt{T})$, showing that the upper bound is tight up to logarithmic factors. Experiments on real-world data demonstrate the effectiveness of our solutions.

Howard H. Qian, Yiting Chen, Gaotian Wang et al.

Successful execution of dexterous robotic manipulation tasks in new environments, such as grasping, depends on the ability to proficiently segment unseen objects from the background and other objects. Previous works in unseen object instance segmentation (UOIS) train models on large-scale datasets, which often leads to overfitting on static visual features. This dependency results in poor generalization performance when confronted with out-of-distribution scenarios. To address this limitation, we rethink the task of UOIS based on the principle that vision is inherently interactive and occurs over time. We propose a novel real-time interactive perception framework, rt-RISeg, that continuously segments unseen objects by robot interactions and analysis of a designed body frame-invariant feature (BFIF). We demonstrate that the relative rotational and linear velocities of randomly sampled body frames, resulting from selected robot interactions, can be used to identify objects without any learned segmentation model. This fully self-contained segmentation pipeline generates and updates object segmentation masks throughout each robot interaction without the need to wait for an action to finish. We showcase the effectiveness of our proposed interactive perception method by achieving an average object segmentation accuracy rate 27.5% greater than state-of-the-art UOIS methods. Furthermore, although rt-RISeg is a standalone framework, we show that the autonomously generated segmentation masks can be used as prompts to vision foundation models for significantly improved performance.

Podshara Chanrungmaneekul, Yiting Chen, Joshua T. Grace et al.

Camera-to-robot (also known as eye-to-hand) calibration is a critical component of vision-based robot manipulation. Traditional marker-based methods often require human intervention for system setup. Furthermore, existing autonomous markerless calibration methods typically rely on pre-trained robot tracking models that impede their application on edge devices and require fine-tuning for novel robot embodiments. To address these limitations, this paper proposes a model-based markerless camera-to-robot calibration framework, ARC-Calib, that is fully autonomous and generalizable across diverse robots and scenarios without requiring extensive data collection or learning. First, exploratory robot motions are introduced to generate easily trackable trajectory-based visual patterns in the camera's image frames. Then, a geometric optimization framework is proposed to exploit the coplanarity and collinearity constraints from the observed motions to iteratively refine the estimated calibration result. Our approach eliminates the need for extra effort in either environmental marker setup or data collection and model training, rendering it highly adaptable across a wide range of real-world autonomous systems. Extensive experiments are conducted in both simulation and the real world to validate its robustness and generalizability.

Sheng Song, Yiting Chen, Duo Xu et al.

Freehand 3D ultrasound enables volumetric imaging by tracking a conventional ultrasound probe during freehand scanning, offering enriched spatial information that improves clinical diagnosis. However, the quality of reconstructed volumes is often compromised by tracking system noise and irregular probe movements, leading to artifacts in the final reconstruction. To address these challenges, we propose ImplicitCell, a novel framework that integrates Implicit Neural Representation (INR) with an ultrasound resolution cell model for joint optimization of volume reconstruction and pose refinement. Three distinct datasets are used for comprehensive validation, including phantom, common carotid artery, and carotid atherosclerosis. Experimental results demonstrate that ImplicitCell significantly reduces reconstruction artifacts and improves volume quality compared to existing methods, particularly in challenging scenarios with noisy tracking data. These improvements enhance the clinical utility of freehand 3D ultrasound by providing more reliable and precise diagnostic information.

Lingyi Hong, Shilin Yan, Renrui Zhang et al.

Visual object tracking aims to localize the target object of each frame based on its initial appearance in the first frame. Depending on the input modility, tracking tasks can be divided into RGB tracking and RGB+X (e.g. RGB+N, and RGB+D) tracking. Despite the different input modalities, the core aspect of tracking is the temporal matching. Based on this common ground, we present a general framework to unify various tracking tasks, termed as OneTracker. OneTracker first performs a large-scale pre-training on a RGB tracker called Foundation Tracker. This pretraining phase equips the Foundation Tracker with a stable ability to estimate the location of the target object. Then we regard other modality information as prompt and build Prompt Tracker upon Foundation Tracker. Through freezing the Foundation Tracker and only adjusting some additional trainable parameters, Prompt Tracker inhibits the strong localization ability from Foundation Tracker and achieves parameter-efficient finetuning on downstream RGB+X tracking tasks. To evaluate the effectiveness of our general framework OneTracker, which is consisted of Foundation Tracker and Prompt Tracker, we conduct extensive experiments on 6 popular tracking tasks across 11 benchmarks and our OneTracker outperforms other models and achieves state-of-the-art performance.

Jie Ren, Die Zhang, Yisen Wang et al.

This paper provides a unified view to explain different adversarial attacks and defense methods, \emph{i.e.} the view of multi-order interactions between input variables of DNNs. Based on the multi-order interaction, we discover that adversarial attacks mainly affect high-order interactions to fool the DNN. Furthermore, we find that the robustness of adversarially trained DNNs comes from category-specific low-order interactions. Our findings provide a potential method to unify adversarial perturbations and robustness, which can explain the existing defense methods in a principle way. Besides, our findings also make a revision of previous inaccurate understanding of the shape bias of adversarially learned features.

Xutian Deng, Yiting Chen, Fei Chen et al.

Medical ultrasound has become a routine examination approach nowadays and is widely adopted for different medical applications, so it is desired to have a robotic ultrasound system to perform the ultrasound scanning autonomously. However, the ultrasound scanning skill is considerably complex, which highly depends on the experience of the ultrasound physician. In this paper, we propose a learning-based approach to learn the robotic ultrasound scanning skills from human demonstrations. First, the robotic ultrasound scanning skill is encapsulated into a high-dimensional multi-modal model, which takes the ultrasound images, the pose/position of the probe and the contact force into account. Second, we leverage the power of imitation learning to train the multi-modal model with the training data collected from the demonstrations of experienced ultrasound physicians. Finally, a post-optimization procedure with guided explorations is proposed to further improve the performance of the learned model. Robotic experiments are conducted to validate the advantages of our proposed framework and the learned models.

Yi Wang, Yuchen He, Xutian Deng et al.

With the potential applications of capsule robots in medical endoscopy, accurate dynamic control of the capsule robot is becoming more and more important. In the scale of a capsule robot, the friction between capsule and the environment plays an essential role in the dynamic model, which is usually difficult to model beforehand. In the paper, a tethered capsule robot system driven by a robot manipulator is built, where a strong magnetic Halbach array is mounted on the robot's end-effector to adjust the state of the capsule. To increase the control accuracy, the friction between capsule and the environment is learned with demonstrated trajectories. With the learned friction model, experimental results demonstrate an improvement of 5.6% in terms of tracking error.

Yiting Chen, Miao Li

The rise of deep learning has greatly transformed the pipeline of robotic grasping from model-based approach to data-driven stream. Along this line, a large scale of grasping data either collected from simulation or from real world examples become extremely important. In this paper, we present our recent work on data generation in simulation for a bin-picking scene. 77 objects from the YCB object data sets are used to generate the dataset with PyBullet, where different environment conditions are taken into account including lighting, camera pose, sensor noise and so on. In all, 100K data samples are collected in terms of ground truth segmentation, RGB, 6D pose and point cloud. All the data examples including the source code are made available online.

Yiting Chen, Chenguang Yang, Miao Li

The ability to segment unknown objects in cluttered scenes has a profound impact on robot grasping. The rise of deep learning has greatly transformed the pipeline of robotic grasping from model-based approach to data-driven stream, which generally requires a large scale of grasping data either collected in simulation or from real-world examples. In this paper, we proposed a sim-to-real framework to transfer the object segmentation model learned in simulation to the real-world. First, data samples are collected in simulation, including RGB, 6D pose, and point cloud. Second, we also present a GAN-based unknown object segmentation method through domain adaptation, which consists of an image translation module and an image segmentation module. The image translation module is used to shorten the reality gap and the segmentation module is responsible for the segmentation mask generation. We used the above method to perform segmentation experiments on unknown objects in a bin-picking scenario. Finally, the experimental result shows that the segmentation model learned in simulation can be used for real-world data segmentation.

Jie Ren, Die Zhang, Yisen Wang et al.

This paper provides a unified view to explain different adversarial attacks and defense methods, i.e. the view of multi-order interactions between input variables of DNNs. Based on the multi-order interaction, we discover that adversarial attacks mainly affect high-order interactions to fool the DNN. Furthermore, we find that the robustness of adversarially trained DNNs comes from category-specific low-order interactions. Our findings provide a potential method to unify adversarial perturbations and robustness, which can explain the existing defense methods in a principle way. Besides, our findings also make a revision of previous inaccurate understanding of the shape bias of adversarially learned features.

Hao Zhang, Xu Cheng, Yiting Chen et al.

In this study, we define interaction components of different orders between two input variables based on game theory. We further prove that interaction components of different orders satisfy several desirable properties.

Hao Zhang, Yiting Chen, Haotian Ma et al.

In order to prevent leaking input information from intermediate-layer features, this paper proposes a method to revise the traditional neural network into the rotation-equivariant neural network (RENN). Compared to the traditional neural network, the RENN uses d-ary vectors/tensors as features, in which each element is a d-ary number. These d-ary features can be rotated (analogous to the rotation of a d-dimensional vector) with a random angle as the encryption process. Input information is hidden in this target phase of d-ary features for attribute obfuscation. Even if attackers have obtained network parameters and intermediate-layer features, they cannot extract input information without knowing the target phase. Hence, the input privacy can be effectively protected by the RENN. Besides, the output accuracy of RENNs only degrades mildly compared to traditional neural networks, and the computational cost is significantly less than the homomorphic encryption.

Hao Zhang, Yiting Chen, Liyao Xiang et al.

We propose a method to revise the neural network to construct the quaternion-valued neural network (QNN), in order to prevent intermediate-layer features from leaking input information. The QNN uses quaternion-valued features, where each element is a quaternion. The QNN hides input information into a random phase of quaternion-valued features. Even if attackers have obtained network parameters and intermediate-layer features, they cannot extract input information without knowing the target phase. In this way, the QNN can effectively protect the input privacy. Besides, the output accuracy of QNNs only degrades mildly compared to traditional neural networks, and the computational cost is much less than other privacy-preserving methods.