arXiv subjects

Arman Roohi

Publications and source records attributed to Arman Roohi.

At least 19 recordsLinked to original sources

arXiv2026

LIMCA: LLM for Automating Analog In-Memory Computing Architecture Design Exploration

Resistive crossbars enabling analog In-Memory Computing (IMC) have emerged as a promising architecture for Deep Neural Network (DNN) acceleration, offering high memory bandwidth and in-situ computation. However, the manual, knowledge-intensive design process and the lack of high-quality circuit netlists have significantly constrained design space exploration and optimization to behavioral system-level tools. In this work, we introduce LIMCA, a novel fine-tune-free Large Language Model (LLM)-driven framework for automating the design and evaluation of IMC crossbar architectures. Unlike traditional approaches, LIMCA employs a No-Human-In-Loop (NHIL) automated pipeline to generate and validate circuit netlists for SPICE simulations, eliminating manual intervention. LIMCA systematically explores the IMC design space by leveraging a structured dataset and LLM-based performance evaluation. Our experimental results on MNIST classification demonstrate that LIMCA successfully generates crossbar designs achieving $\geq$96% accuracy while maintaining a power consumption $\leq$3W, making this the first work in LLM-assisted IMC design space exploration. Compared to existing frameworks, LIMCA provides an automated, scalable, and hardware-aware solution, reducing design exploration time while ensuring user-constrained performance trade-offs.

cs.AR↗

arXiv2026

PICasso: An AI-Enabled Design Framework for Autonomous Optimization of Silicon Photonic Devices

We present PICasso, an AI-assisted framework for automated synthesis, verification, and optimization of photonic integrated circuits (PICs) from natural-language specifications. PICasso couples a structured NL -> YAML -> GDS generation pipeline with PDK aware knowledge injection, automated placement and routing, DRC/LVS validation, and SAX-based photonic simulation. To systematically evaluate AI-driven photonic design, we introduce PIC-Set, a benchmark of 36 parameterized PIC design tasks spanning core photonic primitives and multi-component circuits. Using PIC-Set, we benchmark several state-of-the-art Large Language Models (LLMs) under a unified evaluation protocol, including new metrics such as structural and functional $Spec@k$, optimization efficiency, and robustness under perturbations. Across the benchmark, PICasso significantly improves end-to-end specification satisfaction compared to vanilla LLM generation. Structural $Spec@3$ reaches up to 92.7% and functional $Spec@3$ up to 52% on high-complexity circuits. In addition, PICasso consistently reduces circuit insertion loss, lowering the mean loss from 4.98 dB to 3.25 dB (1.74 dB improvement) through simulation-guided optimization. These results demonstrate that structured domain constraints, physical verification, and simulation feedback transform LLMs from brittle netlist generators into practical PIC design agents capable of producing manufacturable layouts with competitive runtimes relative to manual GUI-based workflows.

cs.AI↗

arXiv2026

Light-Bound Transformers: Hardware-Anchored Robustness for Silicon-Photonic Computer Vision Systems

Deploying Vision Transformers (ViTs) on near-sensor analog accelerators demands training pipelines that are explicitly aligned with device-level noise and energy constraints. We introduce a compact framework for silicon-photonic execution of ViTs that integrates measured hardware noise, robust attention training, and an energy-aware processing flow. We first characterize bank-level noise in microring-resonator (MR) arrays, including fabrication variation, thermal drift, and amplitude noise, and convert these measurements into closed-form, activation-dependent variance proxies for attention logits and feed-forward activations. Using these proxies, we develop Chance-Constrained Training (CCT), which enforces variance-normalized logit margins to bound attention rank flips, and a noise-aware LayerNorm that stabilizes feature statistics without changing the optical schedule. These components yield a practical ``measure $\rightarrow$ model $\rightarrow$ train $\rightarrow$ run'' pipeline that optimizes accuracy under noise while respecting system energy limits. Hardware-in-the-loop experiments with MR photonic banks show that our approach restores near-clean accuracy under realistic noise budgets, with no in-situ learning or additional optical MACs.

cs.ET↗

arXiv2026

GLANCE: Gaze-Led Attention Network for Compressed Edge-inference

Real-time object detection in AR/VR systems faces critical computational constraints, requiring sub-10\,ms latency within tight power budgets. Inspired by biological foveal vision, we propose a two-stage pipeline that combines differentiable weightless neural networks for ultra-efficient gaze estimation with attention-guided region-of-interest object detection. Our approach eliminates arithmetic-intensive operations by performing gaze tracking through memory lookups rather than multiply-accumulate computations, achieving an angular error of $8.32^{\circ}$ with only 393 MACs and 2.2 KiB of memory per frame. Gaze predictions guide selective object detection on attended regions, reducing computational burden by 40-50\% and energy consumption by 65\%. Deployed on the Arduino Nano 33 BLE, our system achieves 48.1\% mAP on COCO (51.8\% on attended objects) while maintaining sub-10\,ms latency, meeting stringent AR/VR requirements by improving the communication time by $\times 177$. Compared to the global YOLOv12n baseline, which achieves 39.2\%, 63.4\%, and 83.1\% accuracy for small, MEDium, and LARGE objects, respectively, the ROI-based method yields 51.3\%, 72.1\%, and 88.1\% under the same settings. This work shows that memory-centric architectures with explicit attention modeling offer better efficiency and accuracy for resource-constrained wearable platforms than uniform processing.

cs.AR↗

arXiv2026

NeuroShield: A Neuro-Symbolic Framework for Adversarial Robustness

Adversarial vulnerability and lack of interpretability are critical limitations of deep neural networks, especially in safety-sensitive settings such as autonomous driving. We introduce \DesignII, a neuro-symbolic framework that integrates symbolic rule supervision into neural networks to enhance both adversarial robustness and explainability. Domain knowledge is encoded as logical constraints over appearance attributes such as shape and color, and enforced through semantic and symbolic logic losses applied during training. Using the GTSRB dataset, we evaluate robustness against FGSM and PGD attacks at a standard $\ell_\infty$ perturbation budget of $\varepsilon = 8/255$. Relative to clean training, standard adversarial training provides modest improvements in robustness ($\sim$10 percentage points). Conversely, our FGSM-Neuro-Symbolic and PGD-Neuro-Symbolic models achieve substantially larger gains, improving adversarial accuracy by 18.1\% and 17.35\% over their corresponding adversarial-training baselines, representing roughly a three-fold larger robustness gain than standard adversarial training provides when both are measured relative to the same clean-training baseline, without reducing clean-sample accuracy. Compared to transformer-based defenses such as LNL-MoEx, which require heavy architectures and extensive data augmentation, our PGD-Neuro-Symbolic variant attains comparable or superior robustness using a ResNet18 backbone trained for 10 epochs. These results show that symbolic reasoning offers an effective path to robust and interpretable AI.

cs.LG↗

arXiv2025

ATM-Net: Adaptive Termination and Multi-Precision Neural Networks for Energy-Harvested Edge Intelligence

ATM-Net is a novel neural network architecture tailored for energy-harvested IoT devices, integrating adaptive termination points with multi-precision computing. It dynamically adjusts computational precision (32/8/4-bit) and network depth based on energy availability via early exit points. An energy-aware task scheduler optimizes the energy-accuracy trade-off. Experiments on CIFAR-10, PlantVillage, and TissueMNIST show ATM-Net achieves up to 96.93% accuracy while reducing power consumption by 87.5% with Q4 quantization compared to 32-bit operations. The power-delay product improves from 13.6J to 0.141J for DenseNet-121 and from 10.3J to 0.106J for ResNet-18, demonstrating its suitability for energy-harvesting systems.

cs.LG↗

arXiv2024

Neuro-Photonix: Enabling Near-Sensor Neuro-Symbolic AI Computing on Silicon Photonics Substrate

Neuro-symbolic Artificial Intelligence (AI) models, blending neural networks with symbolic AI, have facilitated transparent reasoning and context understanding without the need for explicit rule-based programming. However, implementing such models in the Internet of Things (IoT) sensor nodes presents hurdles due to computational constraints and intricacies. In this work, for the first time, we propose a near-sensor neuro-symbolic AI computing accelerator named Neuro-Photonix for vision applications. Neuro-photonix processes neural dynamic computations on analog data while inherently supporting granularity-controllable convolution operations through the efficient use of photonic devices. Additionally, the creation of an innovative, low-cost ADC that works seamlessly with photonic technology removes the necessity for costly ADCs. Moreover, Neuro-Photonix facilitates the generation of HyperDimensional (HD) vectors for HD-based symbolic AI computing. This approach allows the proposed design to substantially diminish the energy consumption and latency of conversion, transmission, and processing within the established cloud-centric architecture and recently designed accelerators. Our device-to-architecture results show that Neuro-Photonix achieves 30 GOPS/W and reduces power consumption by a factor of 20.8 and 4.1 on average on neural dynamics compared to ASIC baselines and photonic accelerators while preserving accuracy.

cs.AR↗

arXiv2024

HiRISE: High-Resolution Image Scaling for Edge ML via In-Sensor Compression and Selective ROI

With the rise of tiny IoT devices powered by machine learning (ML), many researchers have directed their focus toward compressing models to fit on tiny edge devices. Recent works have achieved remarkable success in compressing ML models for object detection and image classification on microcontrollers with small memory, e.g., 512kB SRAM. However, there remain many challenges prohibiting the deployment of ML systems that require high-resolution images. Due to fundamental limits in memory capacity for tiny IoT devices, it may be physically impossible to store large images without external hardware. To this end, we propose a high-resolution image scaling system for edge ML, called HiRISE, which is equipped with selective region-of-interest (ROI) capability leveraging analog in-sensor image scaling. Our methodology not only significantly reduces the peak memory requirements, but also achieves up to 17.7x reduction in data transfer and energy consumption.

cs.CV↗

arXiv2024

Lightator: An Optical Near-Sensor Accelerator with Compressive Acquisition Enabling Versatile Image Processing

This paper proposes a high-performance and energy-efficient optical near-sensor accelerator for vision applications, called Lightator. Harnessing the promising efficiency offered by photonic devices, Lightator features innovative compressive acquisition of input frames and fine-grained convolution operations for low-power and versatile image processing at the edge for the first time. This will substantially diminish the energy consumption and latency of conversion, transmission, and processing within the established cloud-centric architecture as well as recently designed edge accelerators. Our device-to-architecture simulation results show that with favorable accuracy, Lightator achieves 84.4 Kilo FPS/W and reduces power consumption by a factor of ~24x and 73x on average compared with existing photonic accelerators and GPU baseline.

cs.AR↗

arXiv2023

DRAM-Locker: A General-Purpose DRAM Protection Mechanism against Adversarial DNN Weight Attacks

In this work, we propose DRAM-Locker as a robust general-purpose defense mechanism that can protect DRAM against various adversarial Deep Neural Network (DNN) weight attacks affecting data or page tables. DRAM-Locker harnesses the capabilities of in-DRAM swapping combined with a lock-table to prevent attackers from singling out specific DRAM rows to safeguard DNN's weight parameters. Our results indicate that DRAM-Locker can deliver a high level of protection downgrading the performance of targeted weight attacks to a random attack level. Furthermore, the proposed defense mechanism demonstrates no reduction in accuracy when applied to CIFAR-10 and CIFAR-100. Importantly, DRAM-Locker does not necessitate any software retraining or result in extra hardware burden.

cs.AR↗

arXiv2023

Enabling Normally-off In-Situ Computing with a Magneto-Electric FET-based SRAM Design

As an emerging post-CMOS Field Effect Transistor, Magneto-Electric FETs (MEFETs) offer compelling design characteristics for logic and memory applications, such as high-speed switching, low power consumption, and non-volatility. In this paper, for the first time, a non-volatile MEFET-based SRAM design named ME-SRAM is proposed for edge applications which can remarkably save the SRAM static power consumption in the idle state through a fast backup-restore process. To enable normally-off in-situ computing, the ME-SRAM cell is integrated into a novel processing-in-SRAM architecture that exploits a hardware-optimized bit-line computing approach for the execution of Boolean logic operations between operands housed in a memory sub-array within a single clock cycle. Our device-to-architecture evaluation results on Binary convolutional neural network acceleration show the robust performance of ME- SRAM while reducing energy consumption on average by a factor of 5.3 times compared to the best in-SRAM designs.

cs.AR↗

arXiv2023

OISA: Architecting an Optical In-Sensor Accelerator for Efficient Visual Computing

Targeting vision applications at the edge, in this work, we systematically explore and propose a high-performance and energy-efficient Optical In-Sensor Accelerator architecture called OISA for the first time. Taking advantage of the promising efficiency of photonic devices, the OISA intrinsically implements a coarse-grained convolution operation on the input frames in an innovative minimum-conversion fashion in low-bit-width neural networks. Such a design remarkably reduces the power consumption of data conversion, transmission, and processing in the conventional cloud-centric architecture as well as recently-presented edge accelerators. Our device-to-architecture simulation results on various image data-sets demonstrate acceptable accuracy while OISA achieves 6.68 TOp/s/W efficiency. OISA reduces power consumption by a factor of 7.9 and 18.4 on average compared with existing electronic in-/near-sensor and ASIC accelerators.

cs.AR↗

arXiv2023

DIAC: Design Exploration of Intermittent-Aware Computing Realizing Batteryless Systems

Battery-powered IoT devices face challenges like cost, maintenance, and environmental sustainability, prompting the emergence of batteryless energy-harvesting systems that harness ambient sources. However, their intermittent behavior can disrupt program execution and cause data loss, leading to unpredictable outcomes. Despite exhaustive studies employing conventional checkpoint methods and intricate programming paradigms to address these pitfalls, this paper proposes an innovative systematic methodology, namely DIAC. The DIAC synthesis procedure enhances the performance and efficiency of intermittent computing systems, with a focus on maximizing forward progress and minimizing the energy overhead imposed by distinct memory arrays for backup. Then, a finite-state machine is delineated, encapsulating the core operations of an IoT node, sense, compute, transmit, and sleep states. First, we validate the robustness and functionalities of a DIAC-based design in the presence of power disruptions. DIAC is then applied to a wide range of benchmarks, including ISCAS-89, MCNS, and ITC-99. The simulation results substantiate the power-delay-product (PDP) benefits. For example, results for complex MCNC benchmarks indicate a PDP improvement of 61%, 56%, and 38% on average compared to three alternative techniques, evaluated at 45 nm.

cs.AR↗

arXiv2023

Ultra-Fast, High-Performance 8x8 Approximate Multipliers by a New Multicolumn 3,3:2 Inexact Compressor and its Derivatives

A multiplier, as a key component in many different applications, is a time-consuming, energy-intensive computation block. Approximate computing is a practical design paradigm that attempts to improve hardware efficacy while keeping computation quality satisfactory. A novel multicolumn 3,3:2 inexact compressor is presented in this paper. It takes three partial products from two adjacent columns each for rapid partial product reduction. The proposed inexact compressor and its derivates enable us to design a high-speed approximate multiplier. Then, another ultra-fast, high-efficient approximate multiplier is achieved utilizing a systematic truncation strategy. The proposed multipliers accumulate partial products in only two stages, one fewer stage than other approximate multipliers in the literature. Implementation results by Synopsys Design Compiler and 45 nm technology node demonstrates nearly 11.11% higher speed for the second proposed design over the fastest existing approximate multiplier. Furthermore, the new approximate multipliers are applied to the image processing application of image sharpening, and their performance in this application is highly satisfactory. It is shown in this paper that the error pattern of an approximate multiplier, in addition to the mean error distance and error rate, has a direct effect on the outcomes of the image processing application.

cs.AR↗

arXiv2023

NeSe: Near-Sensor Event-Driven Scheme for Low Power Energy Harvesting Sensors

Digital technologies have made it possible to deploy visual sensor nodes capable of detecting motion events in the coverage area cost-effectively. However, background subtraction, as a widely used approach, remains an intractable task due to its inability to achieve competitive accuracy and reduced computation cost simultaneously. In this paper, an effective background subtraction approach, namely NeSe, for tiny energy-harvested sensors is proposed leveraging non-volatile memory (NVM). Using the developed software/hardware method, the accuracy and efficiency of event detection can be adjusted at runtime by changing the precision depending on the application's needs. Due to the near-sensor implementation of background subtraction and NVM usage, the proposed design reduces the data movement overhead while ensuring intermittent resiliency. The background is stored for a specific time interval within NVMs and compared with the next frame. If the power is cut, the background remains unchanged and is updated after the interval passes. Once the moving object is detected, the device switches to the high-powered sensor mode to capture the image.

eess.SP↗

arXiv2022

A Near-Sensor Processing Accelerator for Approximate Local Binary Pattern Networks

In this work, a high-speed and energy-efficient comparator-based Near-Sensor Local Binary Pattern accelerator architecture (NS-LBP) is proposed to execute a novel local binary pattern deep neural network. First, inspired by recent LBP networks, we design an approximate, hardware-oriented, and multiply-accumulate (MAC)-free network named Ap-LBP for efficient feature extraction, further reducing the computation complexity. Then, we develop NS-LBP as a processing-in-SRAM unit and a parallel in-memory LBP algorithm to process images near the sensor in a cache, remarkably reducing the power consumption of data transmission to an off-chip processor. Our circuit-to-application co-simulation results on MNIST and SVHN data-sets demonstrate minor accuracy degradation compared to baseline CNN and LBP-network models, while NS-LBP achieves 1.25 GHz and energy-efficiency of 37.4 TOPS/W. NS-LBP reduces energy consumption by 2.2x and execution time by a factor of 4x compared to the best recent LBP-based networks.

cs.AR↗

arXiv2022

PISA: A Binary-Weight Processing-In-Sensor Accelerator for Edge Image Processing

This work proposes a Processing-In-Sensor Accelerator, namely PISA, as a flexible, energy-efficient, and high-performance solution for real-time and smart image processing in AI devices. PISA intrinsically implements a coarse-grained convolution operation in Binarized-Weight Neural Networks (BWNNs) leveraging a novel compute-pixel with non-volatile weight storage at the sensor side. This remarkably reduces the power consumption of data conversion and transmission to an off-chip processor. The design is completed with a bit-wise near-sensor processing-in-DRAM computing unit to process the remaining network layers. Once the object is detected, PISA switches to typical sensing mode to capture the image for a fine-grained convolution using only the near-sensor processing unit. Our circuit-to-application co-simulation results on a BWNN acceleration demonstrate acceptable accuracy on various image datasets in coarse-grained evaluation compared to baseline BWNN models, while PISA achieves a frame rate of 1000 and efficiency of ~1.74 TOp/s/W. Lastly, PISA substantially reduces data conversion and transmission energy by ~84% compared to a baseline CPU-sensor design.

cs.AR↗

arXiv2020

Entropy-Based Modeling for Estimating Soft Errors Impact on Binarized Neural Network Inference

Over past years, the easy accessibility to the large scale datasets has significantly shifted the paradigm for developing highly accurate prediction models that are driven from Neural Network (NN). These models can be potentially impacted by the radiation-induced transient faults that might lead to the gradual downgrade of the long-running expected NN inference accelerator. The crucial observation from our rigorous vulnerability assessment on the NN inference accelerator demonstrates that the weights and activation functions are unevenly susceptible to both single-event upset (SEU) and multi-bit upset (MBU), especially in the first five layers of our selected convolution neural network. In this paper, we present the relatively-accurate statistical models to delineate the impact of both undertaken SEU and MBU across layers and per each layer of the selected NN. These models can be used for evaluating the error-resiliency magnitude of NN topology before adopting them in the safety-critical applications.

cs.LG↗