arXiv subjects

Zhifei Li

Publications and source records attributed to Zhifei Li.

At least 37 records · Page 2Linked to original sources

arXiv2025

SkyWalker: A Locality-Aware Cross-Region Load Balancer for LLM Inference

Serving Large Language Models (LLMs) efficiently in multi-region setups remains a challenge. Due to cost and GPU availability concerns, providers typically deploy LLMs in multiple regions using instance with long-term commitments, like reserved instances or on-premise clusters, which are often underutilized due to their region-local traffic handling and diurnal traffic variance. In this paper, we introduce SkyWalker, a multi-region load balancer for LLM inference that aggregates regional diurnal patterns through cross-region traffic handling. By doing so, SkyWalker enables providers to reserve instances based on expected global demand, rather than peak demand in each individual region. Meanwhile, SkyWalker preserves KV-Cache locality and load balancing, ensuring cost efficiency without sacrificing performance. SkyWalker achieves this with a cache-aware cross-region traffic handler and a selective pushing based load balancing mechanism. Our evaluation on real-world workloads shows that it achieves 1.12-2.06x higher throughput and 1.74-6.30x lower latency compared to existing load balancers, while reducing total serving cost by 25%.

cs.DC↗

arXiv2025

Structurally Refined Graph Transformer for Multimodal Recommendation

Multimodal recommendation systems utilize various types of information, including images and text, to enhance the effectiveness of recommendations. The key challenge is predicting user purchasing behavior from the available data. Current recommendation models prioritize extracting multimodal information while neglecting the distinction between redundant and valuable data. They also rely heavily on a single semantic framework (e.g., local or global semantics), resulting in an incomplete or biased representation of user preferences, particularly those less expressed in prior interactions. Furthermore, these approaches fail to capture the complex interactions between users and items, limiting the model's ability to meet diverse users. To address these challenges, we present SRGFormer, a structurally optimized multimodal recommendation model. By modifying the transformer for better integration into our model, we capture the overall behavior patterns of users. Then, we enhance structural information by embedding multimodal information into a hypergraph structure to aid in learning the local structures between users and items. Meanwhile, applying self-supervised tasks to user-item collaborative signals enhances the integration of multimodal information, thereby revealing the representational features inherent to the data's modality. Extensive experiments on three public datasets reveal that SRGFormer surpasses previous benchmark models, achieving an average performance improvement of 4.47 percent on the Sports dataset. The code is publicly available online.

cs.IR↗

arXiv2025

Barbarians at the Gate: How AI is Upending Systems Research

Artificial Intelligence (AI) is starting to transform the research process as we know it by automating the discovery of new solutions. Given a task, the typical AI-driven approach is (i) to generate a set of diverse solutions, and then (ii) to verify these solutions and select one that solves the problem. Crucially, this approach assumes the existence of a reliable verifier, i.e., one that can accurately determine whether a solution solves the given problem. We argue that systems research, long focused on designing and evaluating new performance-oriented algorithms, is particularly well-suited for AI-driven solution discovery. This is because system performance problems naturally admit reliable verifiers: solutions are typically implemented in real systems or simulators, and verification reduces to running these software artifacts against predefined workloads and measuring performance. We term this approach as AI-Driven Research for Systems (ADRS), which iteratively generates, evaluates, and refines solutions. Using penEvolve, an existing open-source ADRS instance, we present case studies across diverse domains, including load balancing for multi-region cloud scheduling, Mixture-of-Experts inference, LLM-based SQL queries, and transaction scheduling. In multiple instances, ADRS discovers algorithms that outperform state-of-the-art human designs (e.g., achieving up to 5.0x runtime improvements or 50% cost reductions). We distill best practices for guiding algorithm evolution, from prompt design to evaluator construction, for existing frameworks. We then discuss the broader implications for the systems community: as AI assumes a central role in algorithm design, we argue that human researchers will increasingly focus on problem formulation and strategic guidance. Our results highlight both the disruptive potential and the urgent need to adapt systems research practices in the age of AI.

cs.AI↗

arXiv2025

SCRA-VQA: Summarized Caption-Rerank for Augmented Large Language Models in Visual Question Answering

Acquiring high-quality knowledge is a central focus in Knowledge-Based Visual Question Answering (KB-VQA). Recent methods use large language models (LLMs) as knowledge engines for answering. These methods generally employ image captions as visual text descriptions to assist LLMs in interpreting images. However, the captions frequently include excessive noise irrelevant to the question, and LLMs generally do not comprehend VQA tasks, limiting their reasoning capabilities. To address this issue, we propose the Summarized Caption-Rerank Augmented VQA (SCRA-VQA), which employs a pre-trained visual language model to convert images into captions. Moreover, SCRA-VQA generates contextual examples for the captions while simultaneously summarizing and reordering them to exclude unrelated information. The caption-rerank process enables LLMs to understand the image information and questions better, thus enhancing the model's reasoning ability and task adaptability without expensive end-to-end training. Based on an LLM with 6.7B parameters, SCRA-VQA performs excellently on two challenging knowledge-based VQA datasets: OK-VQA and A-OKVQA, achieving accuracies of 38.8% and 34.6%. Our code is available at https://github.com/HubuKG/SCRA-VQA.

cs.CV↗

arXiv2025

Integrating Object Interaction Self-Attention and GAN-Based Debiasing for Visual Question Answering

Visual Question Answering (VQA) presents a unique challenge by requiring models to understand and reason about visual content to answer questions accurately. Existing VQA models often struggle with biases introduced by the training data, leading to over-reliance on superficial patterns and inadequate generalization to diverse questions and images. This paper presents a novel model, IOG-VQA, which integrates Object Interaction Self-Attention and GAN-Based Debiasing to enhance VQA model performance. The self-attention mechanism allows our model to capture complex interactions between objects within an image, providing a more comprehensive understanding of the visual context. Meanwhile, the GAN-based debiasing framework generates unbiased data distributions, helping the model to learn more robust and generalizable features. By leveraging these two components, IOG-VQA effectively combines visual and textual information to address the inherent biases in VQA datasets. Extensive experiments on the VQA-CP v1 and VQA-CP v2 datasets demonstrate that our model shows excellent performance compared with the existing methods, particularly in handling biased and imbalanced data distributions highlighting the importance of addressing both object interactions and dataset biases in advancing VQA tasks. Our code is available at https://github.com/HubuKG/IOG-VQA.

cs.CV↗

arXiv2025

Deuterium-deuterium fusion charged particle detection using CR-39 and Deep Learning Model

CR-39 solid-state nuclear track detectors are widely used in fusion research for detecting charged particles produced in fusion reactions. However, analyzing increasingly complex and large-scale CR-39 track images to extract meaningful information can be a tedious and time-consuming process, often prone to human errors and bias. To address these challenges, we developed an AI-based classification model capable of differentiating protons, tritons, and helions produced during D-D fusion, using CR-39 track images as input data. The CR-39 track images were processed and used to train a deep learning model. By preprocessing the track images for noise reduction and feature enhancement, we trained the YOLOv8 [1][2] network to distinguish the three particle types with high accuracy. The proposed model achieved a classification accuracy of over 96%, demonstrating its potential for improving automated track analysis in CR-39 detectors. Additionally, the model precisely identifies particle coordinates and counts, enabling comprehensive particle analysis. This study highlights the application of AI in track detection and classification, offering a robust solution for particle identification in CR-39 detector-based experiments.

physics.ins-det↗

arXiv2025

Spark-TTS: An Efficient LLM-Based Text-to-Speech Model with Single-Stream Decoupled Speech Tokens

Recent advancements in large language models (LLMs) have driven significant progress in zero-shot text-to-speech (TTS) synthesis. However, existing foundation models rely on multi-stage processing or complex architectures for predicting multiple codebooks, limiting efficiency and integration flexibility. To overcome these challenges, we introduce Spark-TTS, a novel system powered by BiCodec, a single-stream speech codec that decomposes speech into two complementary token types: low-bitrate semantic tokens for linguistic content and fixed-length global tokens for speaker attributes. This disentangled representation, combined with the Qwen2.5 LLM and a chain-of-thought (CoT) generation approach, enables both coarse-grained control (e.g., gender, speaking style) and fine-grained adjustments (e.g., precise pitch values, speaking rate). To facilitate research in controllable TTS, we introduce VoxBox, a meticulously curated 100,000-hour dataset with comprehensive attribute annotations. Extensive experiments demonstrate that Spark-TTS not only achieves state-of-the-art zero-shot voice cloning but also generates highly customizable voices that surpass the limitations of reference-based synthesis. Source code, pre-trained models, and audio samples are available at https://github.com/SparkAudio/Spark-TTS.

cs.SD↗

arXiv2025

Llasa: Scaling Train-Time and Inference-Time Compute for Llama-based Speech Synthesis

Recent advances in text-based large language models (LLMs), particularly in the GPT series and the o1 model, have demonstrated the effectiveness of scaling both training-time and inference-time compute. However, current state-of-the-art TTS systems leveraging LLMs are often multi-stage, requiring separate models (e.g., diffusion models after LLM), complicating the decision of whether to scale a particular model during training or testing. This work makes the following contributions: First, we explore the scaling of train-time and inference-time compute for speech synthesis. Second, we propose a simple framework Llasa for speech synthesis that employs a single-layer vector quantizer (VQ) codec and a single Transformer architecture to fully align with standard LLMs such as Llama. Our experiments reveal that scaling train-time compute for Llasa consistently improves the naturalness of synthesized speech and enables the generation of more complex and accurate prosody patterns. Furthermore, from the perspective of scaling inference-time compute, we employ speech understanding models as verifiers during the search, finding that scaling inference-time compute shifts the sampling modes toward the preferences of specific verifiers, thereby improving emotional expressiveness, timbre consistency, and content accuracy. In addition, we released the checkpoint and training code for our TTS model (1B, 3B, 8B) and codec model publicly available.

eess.AS↗

arXiv2025

Performance Characteristics of the Battery-Operated Si PIN Diode Detector with Integrated Preamplifier and Data Acquisition Module for Fusion Particle Detection

We present the performance and application of a commercial off-the shelf Si PIN diode (Hamamatsu S14605) as a charged particle detector in a compact ion beam system (IBS) capable of generating D-D and p-B fusion charged particles. This detector is inexpensive, widely available, and operates in photoconductive mode under a reverse bias voltage of 12 V, supplied by an A23 battery. A charge-sensitive preamplifier (CSP) is powered by two 3 V lithium batteries (A123), providing +/-3 V rail voltages. Both the detector and preamplifier circuits are integrated onto the same 4-layer PCB and housed on the vacuum side of the IBS, facing the fusion target. The system employs a CF-2.75 flanged DB-9 connector feedthrough to supply the signal, bias voltage, and rail voltages. To mitigate the high sensitivity of the detector to optical light, a thin aluminum foil assembly is used to block optical emissions from the ion beam and target. Charged particles generate step responses on the preamplifier output, with pulse rise times on the order of 0.2 to 0.3 us. These signals are recorded using a custom-built data acquisition unit, which features an optical fiber data link to ensure electrical isolation of the detector electronics. Subsequent digital signal processing is employed to optimally shape the pulses using a CR-RC^4 filter to produce Gaussian-shaped signals, enabling accurate extraction of energy information. Performance results show that the signal-to-noise ratios (S/N) for D-D fusion charged particles - protons, tritons, and helions - are approximately 30, 10, and 5, respectively, with a shaping time constant of 4 us.

physics.ins-det↗

arXiv2024

CoDiff-VC: A Codec-Assisted Diffusion Model for Zero-shot Voice Conversion

Zero-shot voice conversion (VC) aims to convert the original speaker's timbre to any target speaker while keeping the linguistic content. Current mainstream zero-shot voice conversion approaches depend on pre-trained recognition models to disentangle linguistic content and speaker representation. This results in a timbre residue within the decoupled linguistic content and inadequacies in speaker representation modeling. In this study, we propose CoDiff-VC, an end-to-end framework for zero-shot voice conversion that integrates a speech codec and a diffusion model to produce high-fidelity waveforms. Our approach involves employing a single-codebook codec to separate linguistic content from the source speech. To enhance content disentanglement, we introduce Mix-Style layer normalization (MSLN) to perturb the original timbre. Additionally, we incorporate a multi-scale speaker timbre modeling approach to ensure timbre consistency and improve voice detail similarity. To improve speech quality and speaker similarity, we introduce dual classifier-free guidance, providing both content and timbre guidance during the generation process. Objective and subjective experiments affirm that CoDiff-VC significantly improves speaker similarity, generating natural and higher-quality speech.

cs.SD↗

arXiv2024

Disentangling Heterogeneous Knowledge Concept Embedding for Cognitive Diagnosis on Untested Knowledge

Cognitive diagnosis is a fundamental and critical task in learning assessment, which aims to infer students' proficiency on knowledge concepts from their response logs. Current works assume each knowledge concept will certainly be tested and covered by multiple exercises. However, whether online or offline courses, it's hardly feasible to completely cover all knowledge concepts in several exercises. Restricted tests lead to undiscovered knowledge deficits, especially untested knowledge concepts(UKCs). In this paper, we propose a novel framework for Cognitive Diagnosis called Disentangling Heterogeneous Knowledge Cognitive Diagnosis(DisKCD) on untested knowledge. Specifically, we leverage course grades, exercise questions, and learning resources to learn the potential representations of students, exercises, and knowledge concepts. In particular, knowledge concepts are disentangled into tested and untested based on the limiting actual exercises. We construct a heterogeneous relation graph network via students, exercises, tested knowledge concepts(TKCs), and UKCs. Then, through a hierarchical heterogeneous message-passing mechanism, the fine-grained relations are incorporated into the embeddings of the entities. Finally, the embeddings will be applied to multiple existing cognitive diagnosis models to infer students' proficiency on UKCs. Experimental results on real-world datasets show that the proposed model can effectively improve the performance of the task of diagnosing students' proficiency on UKCs. Our code is available at https://github.com/Hubuers/DisKCD.

cs.AI↗

arXiv2024

Enhancing GNNs with Architecture-Agnostic Graph Transformations: A Systematic Analysis

In recent years, a wide variety of graph neural network (GNN) architectures have emerged, each with its own strengths, weaknesses, and complexities. Various techniques, including rewiring, lifting, and node annotation with centrality values, have been employed as pre-processing steps to enhance GNN performance. However, there are no universally accepted best practices, and the impact of architecture and pre-processing on performance often remains opaque. This study systematically explores the impact of various graph transformations as pre-processing steps on the performance of common GNN architectures across standard datasets. The models are evaluated based on their ability to distinguish non-isomorphic graphs, referred to as expressivity. Our findings reveal that certain transformations, particularly those augmenting node features with centrality measures, consistently improve expressivity. However, these gains come with trade-offs, as methods like graph encoding, while enhancing expressivity, introduce numerical inaccuracies widely-used python packages. Additionally, we observe that these pre-processing techniques are limited when addressing complex tasks involving 3-WL and 4-WL indistinguishable graphs.

cs.LG↗

arXiv2024

Single-Codec: Single-Codebook Speech Codec towards High-Performance Speech Generation

The multi-codebook speech codec enables the application of large language models (LLM) in TTS but bottlenecks efficiency and robustness due to multi-sequence prediction. To avoid this obstacle, we propose Single-Codec, a single-codebook single-sequence codec, which employs a disentangled VQ-VAE to decouple speech into a time-invariant embedding and a phonetically-rich discrete sequence. Furthermore, the encoder is enhanced with 1) contextual modeling with a BLSTM module to exploit the temporal information, 2) a hybrid sampling module to alleviate distortion from upsampling and downsampling, and 3) a resampling module to encourage discrete units to carry more phonetic information. Compared with multi-codebook codecs, e.g., EnCodec and TiCodec, Single-Codec demonstrates higher reconstruction quality with a lower bandwidth of only 304bps. The effectiveness of Single-Code is further validated by LLM-TTS experiments, showing improved naturalness and intelligibility.

eess.AS↗

arXiv2023

PromptSpeaker: Speaker Generation Based on Text Descriptions

Recently, text-guided content generation has received extensive attention. In this work, we explore the possibility of text description-based speaker generation, i.e., using text prompts to control the speaker generation process. Specifically, we propose PromptSpeaker, a text-guided speaker generation system. PromptSpeaker consists of a prompt encoder, a zero-shot VITS, and a Glow model, where the prompt encoder predicts a prior distribution based on the text description and samples from this distribution to obtain a semantic representation. The Glow model subsequently converts the semantic representation into a speaker representation, and the zero-shot VITS finally synthesizes the speaker's voice based on the speaker representation. We verify that PromptSpeaker can generate speakers new from the training set by objective metrics, and the synthetic speaker voice has reasonable subjective matching quality with the speaker prompt.

cs.SD↗

arXiv2023

Distributed Extended Object Tracking Using Coupled Velocity Model from WLS Perspective

This study proposes a coupled velocity model (CVM) that establishes the relation between the orientation and velocity using their correlation, avoiding that the existing extended object tracking (EOT) models treat them as two independent quantities. As a result, CVM detects the mismatch between the prior dynamic model and actual motion pattern to correct the filtering gain, and simultaneously becomes a nonlinear and state-coupled model with multiplicative noise. The study considers CVM to design a feasible distributed weighted least squares (WLS) filter. The WLS criterion requires a linear state-space model containing only additive noise about the estimated state. To meet the requirement, we derive such two separate pseudo-linearized models by using the first-order Taylor series expansion. The separation is merely in form, and the estimates of interested states are embedded as parameters into each other's model, which implies that their interdependency is still preserved in the iterative operation of two linear filters. With the two models, we first propose a centralized WLS filter by converting the measurements from all nodes into a summation form. Then, a distributed consensus scheme, which directly performs an inner iteration on the priors across different nodes, is proposed to incorporate the cross-covariances between nodes. Under the consensus scheme, a distributed WLS filter over a realistic network with ``naive'' node is developed by proper weighting of the priors and measurements. Finally, the performance of proposed filters in terms of accuracy, robustness, and consistency is testified under different prior situations.

eess.SY↗

arXiv2023

Correcting biased noise using Gottesman-Kitaev-Preskill repetition code with noisy ancilla

Concatenation of a bosonic code with a qubit code is one of the promising ways to achieve fault-tolerant quantum computation. As one of the most important bosonic codes, Gottesman-Kitaev-Preskill (GKP) code is proposed to correct small displacement error in phase space. If the noise in phase space is biased, square-lattice GKP code can be concatenated with XZZX surface code or repetition code that promises a high fault-tolerant threshold to suppress the logical error. In this work, we study the performance of GKP repetition codes with physical ancillary GKP qubits in correcting biased noise. We find that there exists a critical value of noise variance for the ancillary GKP qubit such that the logical Pauli error rate decreases when increasing the code size. Furthermore, one round of GKP error correction has to be performed before concatenating with repetition code. Our study paves the way for practical implementation of error correction by concatenating GKP code with low-level qubit codes.

quant-ph↗

arXiv2023

PromptStyle: Controllable Style Transfer for Text-to-Speech with Natural Language Descriptions

Style transfer TTS has shown impressive performance in recent years. However, style control is often restricted to systems built on expressive speech recordings with discrete style categories. In practical situations, users may be interested in transferring style by typing text descriptions of desired styles, without the reference speech in the target style. The text-guided content generation techniques have drawn wide attention recently. In this work, we explore the possibility of controllable style transfer with natural language descriptions. To this end, we propose PromptStyle, a text prompt-guided cross-speaker style transfer system. Specifically, PromptStyle consists of an improved VITS and a cross-modal style encoder. The cross-modal style encoder constructs a shared space of stylistic and semantic representation through a two-stage training process. Experiments show that PromptStyle can achieve proper style transfer with text prompts while maintaining relatively high stability and speaker similarity. Audio samples are available in our demo page.

cs.SD↗

arXiv2022

Distributed Extended Object Tracking Information Filter Over Sensor Networks

This work aims to design a distributed extended object tracking (EOT) system over a realistic network, where both the extent and kinematics are required to retain consensus within the entire network. To this end, we resort to the multiplicative error model (MEM) that allows the extent parameters of perpendicular axis-symmetric objects to have individual uncertainty. To incorporate the MEM into the information filter (IF) style, we use the moment-matching technique to derive two pair linear models with only additive noise. The separation is merely in a fashion, and the cross-correlation between states is preserved as parameters in each other's model. As a result, the closed-form expressions are transferred into an alternating iteration of two linear IFs. With the two models, a centralized IF is proposed wherein the measurements are converted into a summation of innovation parts. Later, under a sensor network with the communication nodes and sensor nodes, we present two distributed IFs through the consensus on information and consensus on measurement schemes, respectively. Moreover, we prove the estimation errors of the proposed filter are exponentially bounded in the mean square. The benefits are testified by numerical experiments in comparison to state-of-the-art filters in literature.

eess.SY↗