Sunnyvale, California

ishihab@iastate.edu

I am a Applied Scientist in Amazon and recently completed my Ph.D. in Computer Science from Iowa State University (October 2025). My research lies at the intersection of artificial intelligence, computer vision, and transportation systems.

Computer Vision Large Language Models Reinforcement Learning Quantum Computing Transportation AI

---

🎓 Research Profile

I specialize in developing AI-driven solutions for safety-critical applications, with expertise spanning computer vision, large language models, reinforcement learning, and quantum computing.

10+

Publications

CVPR, ECAI, EMNLP

Premier Venues

Multiple

Industry Projects

---

📰 Latest News

Feb 21, 2026	Paper accepted at CVPR 2026! Check it out: arxiv.org/pdf/2511.08903
Feb 21, 2026	Serving as Area Chair for the ACL 2026 Industry Track!
Feb 13, 2026	Invited to serve as a reviewer for ACL 2026 Industry Track and ICML 2026!
Jan 31, 2026	Joined Amazon as Applied Scientist II! Excited to start this new chapter.
Jan 15, 2026	Left my position at Dallas College. Grateful for the experiences and growth during my time there!

---

🔬 Research Expertise

Core Technologies

Computer Vision

Crash detection, action localization, traffic analysis

Object Detection Video Analytics Real-time Processing

Large Language Models

Vision-language integration, narrative generation

Vision-LLM Multimodal AI Text Generation

Reinforcement Learning

Autonomous navigation, biomedical optimization

Deep RL Policy Optimization Autonomous Systems

Applied Domains

Transportation AI

ADAS, snowplow navigation, crash prediction

ADAS Traffic Safety IoT DOT

Quantum Computing

QNN architectures, anomaly detection

QNN Network Security Quantum ML

Synthetic Data Generation

CARLA/SUMO simulation, data augmentation

CARLA SUMO Data Synthesis

---

📊 Recent Achievements

4 Papers

CVPR 2026, ECAI 2025

EMNLP 2025 (2 papers)

2025-2026

Iowa DOT

Led AI-based snowplow navigation projects

Crash detection systems

Lead Researcher

Amazon L5

Applied Scientist II

Agentic bot systems

2026 - Present

Amazon

Developed Knowledge Graph frameworks

Novel KGE models

Internship

---

💼 Research Impact

Safety-Critical Systems

Developing AI solutions that directly impact public safety through advanced crash detection and prevention systems

Transportation Innovation

Pioneering autonomous navigation systems and real-time traffic analysis for next-generation transportation

Quantum Security

Advancing network security through quantum neural networks with breakthrough anomaly detection capabilities

---

🤝 Collaboration Opportunities

Let's Work Together

Research Areas

• Computer vision for safety-critical systems
• LLM applications in transportation
• Quantum machine learning
• Reinforcement learning for real-world applications

Opportunities

Research Collaborations Consulting Projects Academic Partnerships Industry Consulting

Email Me More Details

selected publications

1. CVPR

   

   LLM-Guided Probabilistic Fusion for Label-Efficient Document Layout Analysis

   Ibne Farabi Shihab, Sanjeda Akter, and Anuj Sharma

   arXiv preprint arXiv:2511.08903, 2026

   Accepted to IEEE/CVF CVPR 2026

   arXiv

2. 

   HMAE: Self-Supervised Few-Shot Learning for Quantum Spin Systems

   Ibne Farabi Shihab, Sanjeda Akter, and Anuj Sharma

   In ECAI 2025 – 27th European Conference on Artificial Intelligence, 2025

   DOI

3. 

   Efficient Unstructured Pruning of Mamba State-Space Models for Resource-Constrained Environments

   Ibne Farabi Shihab, Sanjeda Akter, and Anuj Sharma

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, Nov 2025

   Abs DOI

   As the deployment of AI models shifts towards edge devices, developing efficient sequence models has become critical. State-space models (SSMs), particularly Mamba, have emerged as strong rivals to Transformers due to their linear-time complexity and impressive performance across a range of tasks. However, their large parameter counts still hinder their use in resource-constrained environments. To address this, we propose a novel unstructured pruning framework specifically tailored for Mamba, achieving up to 70% parameter reduction with only a 3–9% drop in performance. Unlike pruning techniques designed for Transformers, our approach leverages Mamba’s unique recurrent dynamics by incorporating pruning based on both weight and gradient importance to preserve critical parameters, a gradual pruning schedule to maintain model stability, and a global strategy to optimize parameter allocation across the model. Extensive experiments on the WikiText-103, Long Range Arena, and ETT benchmarks demonstrate significant efficiency gains, including 1.77× faster inference and a 46% reduction in memory usage. Our component analysis confirms Mamba’s robustness to pruning, highlighting the framework’s potential for enabling practical deployment while underscoring the need for careful evaluation to avoid introducing biases in sensitive applications.

4. 

   Cache-Efficient Posterior Sampling for Reinforcement Learning with LLM-Derived Priors Across Discrete and Continuous Domains

   Ibne Farabi Shihab, Sanjeda Akter, and Anuj Sharma

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, Nov 2025

   Abs DOI

   Integrating large language models (LLMs) as action proposers in reinforcement learning (RL) significantly boosts performance in text-based environments but incurs prohibitive computational costs. We introduce a cache-efficient framework for Bayesian RL that leverages LLM-derived action suggestions, drastically reducing these costs while maintaining near-optimal performance. Our approach features an adaptive caching mechanism, optimized via meta-learning based on policy performance, to enable efficient inference across text-based games (e.g., TextWorld, ALFWorld) and robotic control tasks (e.g., MuJoCo, MetaWorld). This framework achieves a 3.8×–4.7× reduction in LLM queries and 4.0×–12.0× lower median latencies (85–93ms on consumer hardware), while retaining 96–98% of the uncached policy’s performance. We provide theoretical guarantees on the reliability of cached decisions with Kullback-Leibler (KL) divergence bounds, which are validated empirically by high success rates (90.4–95.6%) in complex text environments. For offline RL, our proposed CQL-Prior variant improves performance by 14–29% and reduces training time by 38–40%. Evaluations across eight diverse tasks demonstrate the framework’s generalizability and practicality for resource-constrained settings, making LLM-guided RL a viable and accessible approach for both text-based and robotic applications.

5. 

   Quantum-driven Zero Trust Architecture with Dynamic Anomaly Detection in 7G Technology: A Neural Network Approach

   Shakil Ahmed, Ibne Farabi Shihab, and Ashfaq Khokhar

   Measurement: Digitalization, Nov 2025
6. CVPR

   

   Deeplocalization: Using Change Point Detection for Temporal Action Localization

   Mohammed Shaiqur Rahman, Ibne Farabi Shihab, Lu Chu, and 1 more author

   In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nov 2024
7. ITSC

   

   Robust and Precise Sidewalk Detection with Ensemble Learning: Enhancing Road Safety and Facilitating Curb Space Management

   Ibne Farabi Shihab, Sudesh Ramesh Bhagat, and Anuj Sharma

   In 2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC), Nov 2023

   DOI

Feel free to reach out for research collaborations or discussions.