About

Peiyang Li, Ph.D., is an Assistant Professor at the Texas A&M AgriLife High Plains Research and Extension Center in Canyon, with an academic appointment in the Department of Biological and Agricultural Engineering at Texas A&M University. His research program focuses on sustainable livestock systems engineering, involving laboratory and field studies on the measurement, mitigation, and modeling of airborne emissions such as gaseous pollutants, particulate matter, bioaerosols, and airborne pathogens. He also works on manure management and utilization, as well as the value-added processing and utilization of co-products from agricultural production. Dr. Li collaborates with multidisciplinary teams including animal scientists, nutritionists, soil scientists, and fellow engineers to conduct solution-oriented research that supports farmers and ranchers. He interacts closely with stakeholders and agricultural producer associations to address industry needs. His goal is to advance the integrity, profitability, and biosecurity of livestock production systems in the Texas Panhandle and Southern High Plains by addressing stakeholder needs, fostering cross-disciplinary collaboration, and mentoring the next generation of engineers and scientists. He is a member of professional societies such as the American Society of Agricultural and Biological Engineers (ASABE) and the Water Environment Federation (WEF).

Research topics

• Artificial Intelligence
• Computer Science
• Algorithm
• Biology

Selected publications

• SpikeX: Exploring Accelerator Architecture and Network-Hardware Co-Optimization for Sparse Spiking Neural Networks

  ArXiv.org · 2025-05-18

  preprintOpen accessSenior author

  Spiking Neural Networks (SNNs) are promising biologically plausible models of computation which utilize a spiking binary activation function similar to that of biological neurons. SNNs are well positioned to process spatiotemporal data, and are advantageous in ultra-low power and real-time processing. Despite a large body of work on conventional artificial neural network accelerators, much less attention has been given to efficient SNN hardware accelerator design. In particular, SNNs exhibit inherent unstructured spatial and temporal firing sparsity, an opportunity yet to be fully explored for great hardware processing efficiency. In this work, we propose a novel systolic-array SNN accelerator architecture, called SpikeX, to take on the challenges and opportunities stemming from unstructured sparsity while taking into account the unique characteristics of spike-based computation. By developing an efficient dataflow targeting expensive multi-bit weight data movements, SpikeX reduces memory access and increases data sharing and hardware utilization for computations spanning across both time and space, thereby significantly improving energy efficiency and inference latency. Furthermore, recognizing the importance of SNN network and hardware co-design, we develop a co-optimization methodology facilitating not only hardware-aware SNN training but also hardware accelerator architecture search, allowing joint network weight parameter optimization and accelerator architectural reconfiguration. This end-to-end network/accelerator co-design approach offers a significant reduction of 15.1x-150.87x in energy-delay-product(EDP) without comprising model accuracy.

  PublisherOA PDFDOI

• Bishop: Sparsified Bundling Spiking Transformers on Heterogeneous Cores with Error-Constrained Pruning

  ArXiv.org · 2025-05-18

  preprintOpen accessSenior author

  We present Bishop, the first dedicated hardware accelerator architecture and HW/SW co-design framework for spiking transformers that optimally represents, manages, and processes spike-based workloads while exploring spatiotemporal sparsity and data reuse. Specifically, we introduce the concept of Token-Time Bundle (TTB), a container that bundles spiking data of a set of tokens over multiple time points. Our heterogeneous accelerator architecture Bishop concurrently processes workload packed in TTBs and explores intra- and inter-bundle multiple-bit weight reuse to significantly reduce memory access. Bishop utilizes a stratifier, a dense core array, and a sparse core array to process MLP blocks and projection layers. The stratifier routes high-density spiking activation workload to the dense core and low-density counterpart to the sparse core, ensuring optimized processing tailored to the given spatiotemporal sparsity level. To further reduce data access and computation, we introduce a novel Bundle Sparsity-Aware (BSA) training pipeline that enhances not only the overall but also structured TTB-level firing sparsity. Moreover, the processing efficiency of self-attention layers is boosted by the proposed Error-Constrained TTB Pruning (ECP), which trims activities in spiking queries, keys, and values both before and after the computation of spiking attention maps with a well-defined error bound. Finally, we design a reconfigurable TTB spiking attention core to efficiently compute spiking attention maps by executing highly simplified "AND" and "Accumulate" operations. On average, Bishop achieves a 5.91x speedup and 6.11x improvement in energy efficiency over previous SNN accelerators, while delivering higher accuracy across multiple datasets.

  PublisherOA PDFDOI

• Serum immunoglobulin G glycosylation profiling in breast cancer using a 56-lectin microarray and subtype-specific patterns

  Gland Surgery · 2025-10-01

  articleOpen access1st author

  Background: Protein glycosylation is the enzymatic addition of sugar chains to specific protein sites and part of this process has a crucial functional modification associated with carcinogenesis and cancer progression. This study aimed to profile and confirm aberrant immunoglobulin G (IgG) protein glycosylation in sera obtained from individuals with breast cancer (BC), with emphasis on breadth of lectin panel and subtype-specific patterns. Methods: A total of 185 sera underwent lectin microarray analysis with 56 lectins. Validation was performed on sera from an additional 20 BC patients and 6 healthy controls. Results: Differential binding intensities of IgG glycans to Agaricus bisporus lectins (ABA) and Salvia sclarea (SSA) were observed between BC patients and healthy controls (P<0.05). In hormone receptor-positive (HR+) vs. HR− BC, lectins Hippeastrum hybrid lectin (HHL), Morniga M (MNA-M), datura stramonium lectin (DSL), and iris hybrid lectin (IRA) showed significantly lower glycan levels (P<0.05). human epidermal growth factor receptor-2-positive (HER2+) vs. HER2-negative (HER2−) BC exhibited significantly higher glycan levels for lectins Jacalin, Artocarpus integrifolia (AIA), MNA-M, HHL, narcissus pseudonarcissus lectin (NPL), Galanthus nivalis lectin (GNL), and soybean agglutinin (SBA) (P<0.05). Verification confirmed increased ABA and SSA in BC sera (P<0.05). SSA, HHL, and NPL recognized glycans in serum IgG from BC subtypes, as confirmed by lectin microarray (P<0.05). This study utilized lectin microarray analysis to profile changes in serum IgG glycosylation patterns in BC. Conclusions: By employing a broad 56-lectin microarray and subtype analysis, our results reveal distinct serum IgG glycosylation profiles associated with different BC molecular markers. These findings offer a novel framework for investigating disease pathogenesis and identifying candidate biomarkers for BC.

  PublisherDOI

• Backpropagation-based learning with local derivative approximation and memory replay in biologically plausible neural systems

  Neurocomputing · 2025-03-07 · 4 citations

  articleOpen accessSenior authorCorresponding

  When learning, the brain modifies individual synaptic connections to reach a desired behavior. Animal and human brains have been shown to be incredibly capable of learning complex and varied functions across a wide variety of tasks. In recent years, artificial neural networks, inspired by human and animal brains, have shown great capabilities in learning a wide variety of difficult tasks. However, artificial neural networks primarily teach themselves through the use of backpropagation, a learning method which has no clear analogue within the brain. Additionally, Artificial Neural Networks primarily use continuous activation functions, which differ significantly from the spiking neuronal behavior present in the brain. In this paper, we discuss and demonstrate a biologically plausible learning method that approximates backpropagation through two techniques on Spiking Neural Networks. First, we show that the local temporal derivatives that are necessary for backpropagation can be approximately recovered through reconstruction using spike timings. Second, we show that through learning during a sleep phase, inspired by neuroscience research into memory replay, the localized parallel feedback path can learn to approximate the derivative through the forward path weight matrix, thus solving the weight transport problem. Finally, we demonstrate that the combination of these two methods can approach or exceed the accuracy of backpropagation-based methods for a variety of neuromorphic vision tasks while maintaining biological plausibility.

  PublisherDOI

• From Learning to Mastery: Achieving Safe and Efficient Real-World Autonomous Driving with Human-In-The-Loop Reinforcement Learning

  ArXiv.org · 2025-10-07

  preprintOpen access

  Autonomous driving with reinforcement learning (RL) has significant potential. However, applying RL in real-world settings remains challenging due to the need for safe, efficient, and robust learning. Incorporating human expertise into the learning process can help overcome these challenges by reducing risky exploration and improving sample efficiency. In this work, we propose a reward-free, active human-in-the-loop learning method called Human-Guided Distributional Soft Actor-Critic (H-DSAC). Our method combines Proxy Value Propagation (PVP) and Distributional Soft Actor-Critic (DSAC) to enable efficient and safe training in real-world environments. The key innovation is the construction of a distributed proxy value function within the DSAC framework. This function encodes human intent by assigning higher expected returns to expert demonstrations and penalizing actions that require human intervention. By extrapolating these labels to unlabeled states, the policy is effectively guided toward expert-like behavior. With a well-designed state space, our method achieves real-world driving policy learning within practical training times. Results from both simulation and real-world experiments demonstrate that our framework enables safe, robust, and sample-efficient learning for autonomous driving.

  PublisherOA PDFDOI

• Reliable Board-Level Degradation Prediction with Monotonic Segmented Regression under Noisy Measurement

  2025-04-28

  articleSenior author

  The increasing complexity of electronic systems in autonomous electric vehicles necessitates robust methods for forecasting the degradation of critical components such as printed circuit boards (PCBs). Various time series forecasting methods have been investigated to predict in-situ resistance degradation under vibration loads. However, these methods failed to capture the degradation trend under strong measurement noise. This paper introduces Monotonic Segmented Linear Regression (MSLR), a novel approach designed to capture monotonic degradation trends in time series data under significant measurement noise. By incorporating monotonic constraints, MSLR effectively models the non-decreasing behavior characteristic of degradation processes. To further enhance reliability of the prediction, we integrate Adaptive Conformal Inference (ACI) with MSLR, enabling the estimation of statistically valid upper bounds for resistance degradation with high confidence. Extensive experiments demonstrate that MSLR outperforms state-of-the-art time series forecasting baselines on real-world PCB degradation datasets.

  PublisherDOI

• Trimming Down Large Spiking Vision Transformers Via Heterogeneous Quantization Search

  2025-07-28 · 1 citations

  articleSenior author

  Spiking neural networks (SNNs) are amenable to deployment on edge devices and neuromorphic hardware due to their lower dissipation. Recently, SNN-based transformers have garnered significant interest, incorporating attention mechanisms akin to their counterparts in artificial neural networks (ANNs) while demonstrating excellent performance. However, deploying large-scale spiking transformer models on resourceconstrained edge devices such as mobile devices still poses significant challenges resulting from the high computational demands of large uncompressed high-precision models. In this work, we introduce a novel heterogeneous quantization method to compress spiking transformers through layer-wise quantization. Our approach optimizes the quantization of each layer using one of two distinct quantization schemes, that is, uniform or power-of-two quantization, with mixed precision. Our heterogeneous quantization demonstrates the feasibility of maintaining high performance for spiking transformers while utilizing an average effective precision of 3.14-3.67 bits with less than a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 \%}$</tex> accuracy drop on neuromorphic DVS Gesture and CIFAR10-DVS datasets. It achieves <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$8.71 \times-10.19 \times$</tex> model compression rate for standard floating-point spiking transformers. Furthermore, the proposed approach achieves substantial energy reductions of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$5.69 \times, 8.72 \times$</tex>, and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$10.2 \times$</tex> on the N-Caltech101, DVS-Gesture, and CIFAR10DVS datasets, respectively, while maintaining high accuracy levels of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$85.3 \%, 97.57 \%$</tex>, and 80.4 %.

  PublisherDOI

• LLM-USO: Large Language Model-Based Universal Sizing Optimizer

  IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · 2025-09-01

  articleSenior author

  The design of analog circuits is a cornerstone of integrated circuit (IC) development, requiring the optimization of complex, interconnected sub-structures such as amplifiers, comparators, and buffers. Traditionally, this process relies heavily on expert human knowledge to refine design objectives by carefully tuning sub-components while accounting for their interdependencies. Existing methods, such as Bayesian Optimization (BO), offer a mathematically driven approach for efficiently navigating large design spaces. However, these methods fall short in two critical areas compared to human expertise: (i) they lack the semantic understanding of the sizing solution space and its direct correlation with design objectives before optimization, and (ii) they fail to reuse knowledge gained from optimizing similar sub-structures across different circuits. To overcome these limitations, we propose the Large Language Model-based Universal Sizing Optimizer (LLM-USO), which introduces a novel method for knowledge representation to encode circuit design knowledge in a structured text format. This representation enables the systematic reuse of optimization insights for circuits with similar sub-structures. LLM-USO employs a hybrid framework that integrates BO with large language models (LLMs) and a learning summary module. This approach serves to: (i) infuse domain-specific knowledge into the BO process and (ii) facilitate knowledge transfer across circuits, mirroring the cognitive strategies of expert designers. Specifically, LLM-USO constructs a knowledge summary mechanism to distill and apply design insights from one circuit to related ones. It also incorporates a knowledge summary critiquing mechanism to ensure the accuracy and quality of the summaries and employs BO-guided suggestion filtering to identify optimal design points efficiently. We evaluate the LLM-USO framework through transfer learning experiments on various analog circuits, demonstrating its ability to improve the quality of circuit design optimization.

  PublisherDOI

• Deep reinforcement learning with evolutionary algorithm-guided imitation for capacitated vehicle routing problems

  Applied Soft Computing · 2025-08-13 · 3 citations

  articleSenior authorCorresponding
  PublisherDOI

• Dynamic prototype-guided structural information maintaining for unsupervised domain adaptation

  Pattern Analysis and Applications · 2025-02-28

  article
  PublisherDOI

Recent grants

• SHF: Small: System-Theoretic Analysis and Design for Dynamic Stability of Memory Devices in Nanoscale CMOS and Beyond

  NSF · $350k · 2009–2013

• Collaborative Research: SHF: Medium: Data-Efficient Uncovering of Rare Design Failures for Reliability-Critical Circuits

  NSF · $633k · 2020–2026

• Thermal-Aware GPU-Based Design Engine for On-Chip Power Delivery in Power-Efficient Multi-Core Chips

  NSF · $255k · 2009–2013

• FET: Small: Heterogeneous Learning Architectures and Training Algorithms for Hardware Accelerated Deep Spiking Neural Computation

  NSF · $499k · 2019–2024

• CAREER: Parallel CAD Algorithms on Emerging Multi-Core Platforms

  NSF · $400k · 2008–2014

Frequent coauthors

• Guoyao Wu

  Texas A&M University

  16 shared

• Zhuo Feng

  16 shared

• Wenrui Zhang

  16 shared

• Yongtae Kim

  Kyungpook National University

  11 shared

• Malachi Schram

  9 shared

• Karthik Somayaji NS

  University of California, Santa Barbara

  9 shared

• Xiaoji Ye

  Guangdong Academy of Sciences

  9 shared

• Wengu Chen

  Institute of Applied Physics and Computational Mathematics

  9 shared

Education

• B.S., Biological Systems Engineering

  Iowa State University

• B.S., Environmental Science

  Iowa State University

• M.S., Agricultural and Biosystems Engineering

  Iowa State University

• Ph.D., Agricultural and Biosystems Engineering

  Iowa State University