arXiv subjects

Boyu Li

Publications and source records attributed to Boyu Li.

At least 19 recordsLinked to original sources

arXiv2026

Beacon: LLM Multi-Agent Driven Hardware Design Space Exploration for Heterogeneous Multi-Chiplet Deep Learning Accelerators

Heterogeneous multi-chiplet accelerators allow chiplets to be configured independently to better match different operator characteristics and improve inference efficiency. However, heterogeneity makes simulator evaluation expensive, limiting the number of iterations affordable for hardware design space exploration (HW-DSE). Mainstream data-driven methods rely mainly on final metrics and a few predefined states, and require many search iterations to implicitly learn the relationships between input parameters and optimization objectives, making them less effective in this setting. In practice, evaluators also generate detailed reports on execution timelines, resource utilization, memory accesses, and communication behavior. Large language models (LLMs) can combine domain knowledge with these reports to explicitly identify bottleneck locations, degradation causes, and parameter adjustment directions, thereby improving each design decision under limited iteration budgets. Based on this observation, we propose Beacon, a report-driven LLM multi-agent framework for heterogeneous multi-chiplet HW-DSE. Beacon employs hierarchical agents for bottleneck localization, root-cause diagnosis, and hardware candidate generation, together with an Analysis Toolbox and RAG memory for closed-loop search. Under the same limited iteration budget, Beacon reduces the composite latency-energy-monetary-cost objective by 25.1\%--93.5\% compared with random search, Bayesian optimization, and reinforcement learning.

cs.AR↗

arXiv2026

Hybrid-LUT: Channel-Aware Hybrid Lookup Table and Filtering for Efficient Image Denoising

Lookup table (LUT)-based image denoising methods have attracted increasing attention due to their high efficiency and hardware-friendly properties. However, existing RGB-LUT approaches require three identical LUTs to process RGB channels in parallel, resulting in large on-chip SRAM consumption. A simple alternative is to apply LUT processing only to the luminance (Y) channel in the YUV color space to reduce memory usage. However, this naive strategy leads to degraded restoration quality, since ignoring the chrominance (UV) channels introduces color distortion and residual artifacts. In this work, we propose Hybrid-LUT, a YUV-based asymmetric channel-processing framework that combines LUT and filtering in a unified design. Specifically, a multi-band LUT branch with pixel-level weight fusion is applied to the Y channel to recover fine textures, while lightweight filtering is used for the UV channels to maintain color consistency. This design reduces LUT storage by two-thirds compared with RGB-LUT methods while maintaining the same runtime throughput. Extensive experiments show that Hybrid-LUT achieves state-of-the-art (SOTA) performance across multiple benchmarks with only 421 KB of storage. In particular, our method surpasses existing LUT-based denoising approaches by at least 0.63 dB CPSNR on real-world datasets, demonstrating its effectiveness for image denoising on resource-constrained edge devices. The project is available at https://github.com/Ai-ZL/Hybrid-LUT .

cs.CV↗

arXiv2026

Spatula: Exploring On-Demand In-Situ Interfaces and Interaction for Attribute Control

Controlling attributes is a critical step toward achieving the final creative outcome, yet current approaches fall short in supporting users in the iterative refinement of generative content. We propose Spatula, a proof-of-concept system that generates on-demand, in-situ attribute control interfaces and interactions for creating motion graphics. Building on a technical probe that automatically analyzes animation context and generates corresponding attributes and UI, we frame attribute control as an explorable landscape and explore the attribute control space along four key dimensions: Discoverability, Resolution, Scope, and Expandability. Findings from a user study (N=12) show that our system provides intuitive and convenient interactions while supporting diverse needs for fine-grained parameter control. Furthermore, our applications demonstrate that the plug-and-play design generalizes to other domains, such as web design and 3D modeling.

cs.HC↗

arXiv2026

AV-SyncBench: Decoupled Benchmarking of Temporal and Semantic Audio-Visual Synchronization

Audio-visual feature extraction is a fundamental component of multimodal understanding and generation tasks. However, existing evaluation protocols for feature extraction models exhibit dimensional bias, typically focusing on either semantic matching or temporal offset detection. Moreover, their data construction remains coupled, preventing independent assessment of temporal and semantic consistency. We propose AV-SyncBench, the first benchmark to fully separate temporal and semantic evaluation for audio-visual synchronization. Built from in-the-wild videos, it spans Voice, Music, and Sound across 10 scenarios and 5 challenge tasks. Data are automatically filtered and manually verified to ensure on-screen sound sources. The benchmark contains 3,269 videos and 38,390 samples, and we evaluate five representative models to quantify feature quality for alignment and downstream tasks. The code and dataset are available at: https://fgt7t6g.github.io/AV-SyncBench.

cs.CV↗

arXiv2026

Mutation and crossover of simplicial complexes

Color graphs and their subgraphs, referred to as bubble graphs, correspond bijectively to the simplicial complexes of pseudomanifolds and their subsimplices, respectively. In this paper, we introduce matrix representations for colored graphs and their associated bubble graphs. By using this correspondence, we define simplicial-complex matrices and subsimplex matrices that encode the simplicial complexes of pseudomanifolds and their subsimplices. Moreover, we formulate mutation and crossover operations on colored graphs. Through the established correspondence among simplicial complexes, colored graphs, and simplicial-complex matrices, we extend these operations to simplicial complexes and simplicial-complex matrices. We further implement an algorithm generating simplicial-complex matrices and a genetic algorithm performing mutation and crossover of them to produce pseudomanifolds exhibiting diverse topologies. In addition, we implement procedures for decomposing the generated simplicial-complex matrices into simplex matrices, reconstructing the simplicial complexes of the associated pseudomanifolds from this information, and computing geometric quantities such as the volume, circumcenter, and dual-simplex volume of each simplex.

hep-th↗

arXiv2026

FreeSonic: Training-Free Temporal-Aware Decoupled Attention for Precise Audio Editing

Text-to-audio (TTA) generation has made significant strides, yet achieving precise and consistent audio editing remains a major challenge. However, existing methods struggle to balance temporal consistency with background preservation. In this paper, we propose FreeSonic, a training-free framework leveraging the state-of-the-art Rectified Flow-based TangoFlux model. FreeSonic utilizes an optimized inversion-reverse process and joint text-audio attention maps for precise target segment extraction. For content editing, a novel scheduled attention decoupling confines modifications to target regions while preserving original acoustic context. Furthermore, task-oriented noise injection enhances versatility for tasks such as audio removal and non-rigid replacement. Extensive experimental results demonstrate that FreeSonic achieves a superior balance by providing a high-fidelity and efficient solution for precise and consistent audio editing. Project and demos: https://free-sonic.github.io/

cs.SD↗

arXiv2026

VoiceGiraffe: A Benchmark for Extreme Long-Context Audio-Language Understanding

While large audio language models (LALMs) have achieved remarkable progress in audio processing at the second- or minute-level scale, understanding hour-level audio remains a fundamental bottleneck. Existing benchmarks predominantly rely on short clips or artificially concatenated segments, failing to faithfully assess LALM capacity for long-range information comprehension in real-world scenarios such as podcasts and lengthy speeches. To address this gap, we introduce VoiceGiraffe, a novel benchmark designed to rigorously evaluate LALMs across diverse real-world scenarios, modalities, and languages under long-context settings. It comprises 1500 curated triplets structured into a dual-level taxonomy of single-hop perception and multi-hop reasoning. We evaluate a broad suite of open-source and proprietary LALMs against human performance. Results underscore three fundamental findings. First, VoiceGiraffe remains highly challenging and far from saturation. Second, we show that no single inference paradigm universally dominates. The E2E inference benefits models with native long-context audio understanding, cascaded caption aggregation stabilizes small models overwhelmed by hour-scale audio, and reasoning-enhanced cascading with external LLM helps weaker models but can bottleneck stronger proprietary systems. Third, we reveal long-range memory persistence as a key bottleneck. LALMs are better at answering questions that require connecting salient causal cues than those requiring sustained tracking of sparse events across long audio, whereas humans show the opposite pattern. These findings position VoiceGiraffe as a challenging and diagnostic testbed for long-form audio understanding, highlighting the need for LALMs with persistent memory and robust long-range aggregation.

cs.SD↗

arXiv2026

X-DiffVLA: X-Embodied Diffusion Action Heads for Vision-Language-Action Models

Learning universal policies from cross-embodied data remains a fundamental challenge in robotics. Although Vision-Language-Action (VLA) models are pre-trained on large and diverse datasets, they typically rely on embodiment-specific fine-tuning to achieve strong performance in downstream tasks. This requirement severely limits their generalization capability and restricts knowledge transfer across embodiments performing similar tasks. To overcome these limitations, we focus on cross-embodied settings with shared robotic bases and heterogeneous end-effectors, and propose X-DiffVLA, a diffusion-based VLA model featuring a unified cross-embodied action head. X-DiffVLA can leverage the generative strengths of diffusion models to capture both the diversity and latent correlations in cross-embodied datasets. Specifically, we introduce Embodiment Forcing, a classifier-free guidance technique to implicitly steer action generation toward embodiment-specific functional components, capturing fine-grained structural nuances without explicit supervision. In addition, a Morphological Tree Diffusion approach is designed to strengthen behavioral correlations across diverse end-effectors, maximizing the transferability of heterogeneous demonstrations. Experimental results across RoboCasa and Isaac Gym, covering different embodiments from grippers to dexterous hands, show that X-DiffVLA achieves state-of-the-art performance, with improvements of 15.3% and 12.5%, respectively. Real-world evaluations further validate the robustness of the proposed framework and its effectiveness in scalable cross-embodied policy learning.

cs.RO↗

arXiv2026

NeuroSleep: Neuromorphic Event-Driven Single-Channel EEG Sleep Staging for Edge-Efficient Sensing

Objective. Reliable, continuous neural sensing on wearable edge platforms is fundamental to long-term health monitoring; however, for electroencephalography (EEG)-based sleep monitoring, dense high-frequency processing is often computationally prohibitive under tight energy budgets. Approach. To address this bottleneck, this paper proposes NeuroSleep, an integrated event-driven sensing and inference system for energy-efficient sleep staging. NeuroSleep first converts raw EEG into complementary multi-scale bipolar event streams using Residual Adaptive Multi-Scale Delta Modulation (R-AMSDM), enabling an explicit fidelity-sparsity trade-off at the sensing front end. Furthermore, NeuroSleep adopts a hierarchical inference architecture that comprises an Event-based Adaptive Multi-scale Response (EAMR) module for local feature extraction, a Local Temporal-Attention Module (LTAM) for context aggregation, and an Epoch-Leaky Integrate-and-Fire (ELIF) module to capture long-term state persistence. Main results. Experimental results using subject-independent 5-fold cross-validation on the Sleep-EDF Expanded sleep-cassette (SC) subset with single-channel EEG demonstrate that NeuroSleep achieves a mean accuracy of 74.2% with only 0.932 M parameters while reducing sparsity-adjusted effective operations by approximately 53.6% relative to dense processing. Compared to the representative dense Transformer baseline, NeuroSleep improves accuracy by 7.5% with a 45.8% reduction in computational load. Significance. By coupling neuromorphic event encoding with state-aware context modeling, NeuroSleep offers a deployment-oriented framework for single-channel sleep staging that reduces redundant high-rate processing and improves energy scalability for wearable and edge platforms.

eess.SP↗

arXiv2026

Direct vs. Score-based Selection: Understanding the Heisenberg Effect in Target Acquisition Across Input Modalities in Virtual Reality

Target selection is a fundamental interaction in virtual reality (VR). But the act of confirming a selection, such as a button press or pinch, can disturb the tracked pose and shift the intended target, which is referred to as the Heisenberg Effect. Prior research has mainly investigated controller input. However, it remains unclear how the effect manifests in the bare-hand input and how score-based techniques may mitigate the effect in different spatial variations. To fill the gap, we conduct a within-subject study to examine the Heisenberg Effect across two input modalities (i.e., controller and hand) and two selection mechanisms (i.e., direct and score-based). Our results show that hand input is more susceptible to the Heisenberg Effect, with direct selection more influenced by target width and score-based selection more sensitive to target density. Based on previous vote-oriented technique and our temporal analysis, we introduce weighted VOTE, a history-based intention accuracy model for target voting, that reweights recent interaction intent to counteract input disturbances. Our evaluation shows the method improves selection accuracy compared to baseline techniques. Finally, we discuss future directions for adaptive selection methods.

cs.HC↗

arXiv2026

Real-Time Human Activity Recognition on Edge Microcontrollers: Dynamic Hierarchical Inference with Multi-Spectral Sensor Fusion

The demand for accurate on-device pattern recognition in edge applications is intensifying, yet existing approaches struggle to reconcile accuracy with computational constraints. To address this challenge, a resource-aware hierarchical network based on multi-spectral fusion and interpretable modules, namely the Hierarchical Parallel Pseudo-image Enhancement Fusion Network (HPPI-Net), is proposed for real-time, on-device Human Activity Recognition (HAR). Deployed on an ARM Cortex-M4 microcontroller for low-power real-time inference, HPPI-Net achieves 96.70% accuracy while utilizing only 22.3 KiB of RAM and 439.5 KiB of ROM after optimization. HPPI-Net employs a two-layer architecture. The first layer extracts preliminary features using Fast Fourier Transform (FFT) spectrograms, while the second layer selectively activates either a dedicated module for stationary activity recognition or a parallel LSTM-MobileNet network (PLMN) for dynamic states. PLMN fuses FFT, Wavelet, and Gabor spectrograms through three parallel LSTM encoders and refines the concatenated features using Efficient Channel Attention (ECA) and Depthwise Separable Convolution (DSC), thereby offering channel-level interpretability while substantially reducing multiply-accumulate operations. Compared with MobileNetV3, HPPI-Net improves accuracy by 1.22% and reduces RAM usage by 71.2% and ROM usage by 42.1%. These results demonstrate that HPPI-Net achieves a favorable accuracy-efficiency trade-off and provides explainable predictions, establishing a practical solution for wearable, industrial, and smart home HAR on memory-constrained edge platforms.

cs.CV↗

arXiv2026

SketchDynamics: Exploring Free-Form Sketches for Dynamic Intent Expression in Animation Generation

Sketching provides an intuitive way to convey dynamic intent in animation authoring (i.e., how elements change over time and space), making it a natural medium for automatic content creation. Yet existing approaches often constrain sketches to fixed command tokens or predefined visual forms, overlooking their freeform nature and the central role of humans in shaping intention. To address this, we introduce an interaction paradigm where users convey dynamic intent to a vision-language model via free-form sketching, instantiated here in a sketch storyboard to motion graphics workflow. We implement an interface and improve it through a three-stage study with 24 participants. The study shows how sketches convey motion with minimal input, how their inherent ambiguity requires users to be involved for clarification, and how sketches can visually guide video refinement. Our findings reveal the potential of sketch and AI interaction to bridge the gap between intention and outcome, and demonstrate its applicability to 3D animation and video generation.

cs.HC↗

arXiv2026

Soft Responsive Materials Enhance Humanoid Safety

Humanoid robots are envisioned as general-purpose platforms in human-centered environments, yet their deployment is limited by vulnerability to falls and the risks posed by rigid metal-plastic structures to people and surroundings. We introduce a soft-rigid co-design framework that leverages non-Newtonian fluid-based soft responsive materials to enhance humanoid safety. The material remains compliant during normal interaction but rapidly stiffens under impact, absorbing and dissipating fall-induced forces. Physics-based simulations guide protector placement and thickness and enable learning of active fall policies. Applied to a 42 kg life-size humanoid, the protector markedly reduces peak impact and allows repeated falls without hardware damage, including drops from 3 m and tumbles down long staircases. Across diverse scenarios, the approach improves robot robustness and environmental safety. By uniting responsive materials, structural co-design, and learning-based control, this work advances interact-safe, industry-ready humanoid robots.

cs.RO↗

arXiv2025

An Ando-type dilation on right LCM monoids

We establish an Ando-type dilation theorem for a pair of commuting contractions together with a representation of a right LCM monoid via either the Cartesian or the free product. We prove that if each individual contraction together with the monoid representation has $*$-regular dilation, then they can be dilated to commuting isometries and an isometric representation of the monoid. This extends an earlier result of Barik and Das.

math.OA↗

arXiv2025

Reconstruction as a Bridge for Event-Based Visual Question Answering

Integrating event cameras with Multimodal Large Language Models (MLLMs) promises general scene understanding in challenging visual conditions, yet requires navigating a trade-off between preserving the unique advantages of event data and ensuring compatibility with frame-based models. We address this challenge by using reconstruction as a bridge, proposing a straightforward Frame-based Reconstruction and Tokenization (FRT) method and designing an efficient Adaptive Reconstruction and Tokenization (ART) method that leverages event sparsity. For robust evaluation, we introduce EvQA, the first objective, real-world benchmark for event-based MLLMs, comprising 1,000 event-Q&A pairs from 22 public datasets. Our experiments demonstrate that our methods achieve state-of-the-art performance on EvQA, highlighting the significant potential of MLLMs in event-based vision.

cs.CV↗

arXiv2025

Compass: Co-Exploration of Mapping and Hardware for Heterogeneous Multi-Chiplet Accelerators Targeting LLM Inference Service Workloads

Large language models (LLMs) bring huge computational demands, which makes multi-chiplet accelerators that can integrate large-scale computing resources a powerful solution. However, existing design space exploration (DSE) efforts for such accelerators primarily focus on traditional CNN/Transformer workloads and fall short in supporting the highly dynamic behavior of real-world LLM inference services. This dynamic nature manifests in two key aspects: 1) Mixed request types: the prefill and decode phases exhibit significantly different computational patterns and are frequently interleaved by modern system-level service schedulers; 2) Variable sequence lengths: the sequence length differences across requests can span several orders of magnitude, rendering padding-based assumptions inefficient. Moreover, many prior works assume homogeneous chiplets and overlook the potential beneficial interaction between LLM dynamics and heterogeneous chiplet architectures. To bridge this gap, we introduce Compass, a co-exploration framework designed to optimize mapping strategies and hardware design for multi-chiplet accelerators, specifically tailored for dynamic LLM workloads. First, we propose a computation execution graph-based mapping encoding scheme that decouples micro-batch and layer dimensions, enabling fine-grained execution control on heterogeneous chiplets and flexibly representing various parallelism strategies. Second, based on this scheme, we develop the Compass framework itself, which integrates an evaluation engine, a mapping generation engine based on genetic algorithm, and a hardware sampling engine based on Bayesian optimization, enabling fast and flexible cross-level co-design. Compared with the SOTA DSE works Gemini and MOHaM, Compass reduces latency by 63.92\% and energy by 40.32\% on average in various scenarios, with only a 3.11\% increase in monetary cost.

cs.AR↗

arXiv2025

Time-Layer Adaptive Alignment for Speaker Similarity in Flow-Matching Based Zero-Shot TTS

Flow-Matching (FM)-based zero-shot text-to-speech (TTS) systems exhibit high-quality speech synthesis and robust generalization capabilities. However, the speaker representation ability of such systems remains underexplored, primarily due to the lack of explicit speaker-specific supervision in the FM framework. To this end, we conduct an empirical analysis of speaker information distribution and reveal its non-uniform allocation across time steps and network layers, underscoring the need for adaptive speaker alignment. Accordingly, we propose Time-Layer Adaptive Speaker Alignment (TLA-SA), a strategy that enhances speaker consistency by jointly leveraging temporal and hierarchical variations. Experimental results show that TLA-SA substantially improves speaker similarity over baseline systems on both research- and industrial-scale datasets and generalizes well across diverse model architectures, including decoder-only language model (LM)-based and free TTS systems. A demo is provided.

eess.AS↗

arXiv2025

RoboGPT-R1: Enhancing Robot Task Planning with Reinforcement Learning

Improving the reasoning capabilities of embodied agents is crucial for robots to complete complex human instructions in long-view manipulation tasks successfully. Despite the success of large language models and vision language models based on Supervised Fine-Tuning (SFT) in planning tasks, they continue facing challenges in performing long-horizon manipulation tasks in complex real-world environments, owing to their restricted common sense and reasoning capabilities. Considering that aligning general-purpose vision language models to robotic planning tasks via supervised fine-tuning suffers from poor generalization and insufficient physical understanding, we propose RoboGPT-R1, a two-stage fine-tuning framework for embodied planning. In this framework, supervised training acquires foundational knowledge through expert sequences, followed by RL to address the model's shortcomings in visual-spatial understanding and reasoning. To achieve physical understanding and action sequence consistency in multi-step reasoning tasks, we design a rule-based reward function that simultaneously considers long-horizon performance and action constraint in the environment. The reasoning model, trained on Qwen2.5-VL-3B, significantly outperforms the larger-scale model, GPT-4o-mini, by 21.33% and surpasses other work trained on Qwen2.5-VL-7B by 20.33% on the EmbodiedBench benchmark.

cs.AI↗