arXiv subjects

Mehrzad Mohammadi

Publications and source records attributed to Mehrzad Mohammadi.

4 recordsLinked to original sources

arXiv2025

A Multi-Modal Neuro-Symbolic Approach for Spatial Reasoning-Based Visual Grounding in Robotics

Visual reasoning, particularly spatial reasoning, is a challenging cognitive task that requires understanding object relationships and their interactions within complex environments, especially in robotics domain. Existing vision_language models (VLMs) excel at perception tasks but struggle with fine-grained spatial reasoning due to their implicit, correlation-driven reasoning and reliance solely on images. We propose a novel neuro_symbolic framework that integrates both panoramic-image and 3D point cloud information, combining neural perception with symbolic reasoning to explicitly model spatial and logical relationships. Our framework consists of a perception module for detecting entities and extracting attributes, and a reasoning module that constructs a structured scene graph to support precise, interpretable queries. Evaluated on the JRDB-Reasoning dataset, our approach demonstrates superior performance and reliability in crowded, human_built environments while maintaining a lightweight design suitable for robotics and embodied AI applications.

cs.RO↗

arXiv2025

JRDB-Reasoning: A Difficulty-Graded Benchmark for Visual Reasoning in Robotics

Recent advances in Vision-Language Models (VLMs) and large language models (LLMs) have greatly enhanced visual reasoning, a key capability for embodied AI agents like robots. However, existing visual reasoning benchmarks often suffer from several limitations: they lack a clear definition of reasoning complexity, offer have no control to generate questions over varying difficulty and task customization, and fail to provide structured, step-by-step reasoning annotations (workflows). To bridge these gaps, we formalize reasoning complexity, introduce an adaptive query engine that generates customizable questions of varying complexity with detailed intermediate annotations, and extend the JRDB dataset with human-object interaction and geometric relationship annotations to create JRDB-Reasoning, a benchmark tailored for visual reasoning in human-crowded environments. Our engine and benchmark enable fine-grained evaluation of visual reasoning frameworks and dynamic assessment of visual-language models across reasoning levels.

cs.CV↗

arXiv2024

A Scalable Multi-Layered Blockchain Architecture for Enhanced EHR Sharing and Drug Supply Chain Management

In recent years, the healthcare sector's transition to digital platforms has intensified concerns over data security, privacy, and scalability. Blockchain technology offers a decentralized, secure, and immutable solution to these challenges. This paper presents a scalable, multi-layered blockchain architecture for secure Electronic Health Record (EHR) sharing and drug supply chain management. The proposed framework introduces five distinct layers that enhance system performance, security, and patient-centric access control. By implementing parallelism, the system significantly increases transaction throughput and reduces network traffic. Our solution ensures data integrity, privacy, and interoperability, making it compatible with existing healthcare systems. Experimental results, conducted using the Caliper benchmark, demonstrate notable improvements in transaction throughput and reduced communication overhead. Additionally, the framework provides transparency and real-time drug supply chain monitoring, empowering decision-makers with critical insights.

cs.CR↗

arXiv2024

SCALHEALTH: Scalable Blockchain Integration for Secure IoT Healthcare Systems

Internet of Things (IoT) devices are capable of allowing for far-reaching access to and evaluation of patient data to monitor health and diagnose from a distance. An electronic healthcare system that checks patient data, prepares medicines and provides financial assistance is necessary. Providing safe data transmission, monitoring, decentralization, preserving patient privacy, and maintaining confidentiality are essential to an electronic healthcare system. In this study, we introduce (SCALHEALTH) which is a blockchain-based scheme of the Hyperledger Fabric consortium. In this study, we use authentication to agree on a common key for data encryption to send data confidentially. Also, sending data through IPFS is decentralized. Non-fungible token (NFT) is used to send patient prescriptions to pharmacies and insurance companies to ensure the authenticity of patient prescriptions. As the system's main body, blockchain creates authorization and validation for all devices and institutions. Also, all metadata in the system is recorded on the blockchain to maintain integrity, transparency, and timely data monitoring. The proposed study uses two types of blockchain: a health blockchain and a financial blockchain. The financial blockchain is for financial transactions and is based on Ethereum. The health blockchain also introduces a mechanism that allows several blockchains to be active in parallel, instead of only one blockchain. The prototype of this mechanism is simulated in two scenarios. In comparison to the normal state, the proposed plan has superior results.

cs.CR↗