About Me

I am a researcher in efficient 3D deep learning and geometric AI for robotic perception, with interests in point cloud understanding, geometric representation learning, and resource-efficient deployment on edge devices.

My recent work has resulted in first-author publications at NeurIPS 2026, ICRA 2026 and IEEE IV 2026 (Oral).

Research

Efficient 3D Deep Learning

Designing lightweight point cloud and 3D vision models that achieve strong performance with low computational cost.

Geometric Representation Learning

Learning structured representations of 3D geometry for recognition, segmentation, and scene understanding.

3D Reconstruction and Perception

Exploring 3D representations for reconstruction, scene understanding, and robotic perception.

Model Compression & Edge Deployment

Optimizing deep learning models for real-time inference, memory efficiency, and deployment on edge devices.

Publications

• Relational Feature Distillation (NeurIPS 2026) — Relational distillation for efficient 3D point cloud segmentation
• SLNet (ICRA 2026) — Super-lightweight geometry-adaptive network for efficient 3D point cloud recognition
• NPNet (IEEE IV 2026, Oral) — Non-parametric 3D learning with adaptive Gaussian–Fourier encoding
• ModelNet-R / Point-SkipNet (IPRIA 2025, Oral) — Dataset refinement and efficient graph-based architecture
• AniGeo (In Preparation) — Adaptive anisotropic geometric embeddings for efficient 3D recognition

Experience

Independent Research Collaboration (2024–Present) Conducting research on efficient 3D point cloud learning, lightweight neural architectures, CUDA-aware optimization, and deployment benchmarking on NVIDIA Jetson platforms. This work has led to multiple first-author publications at ICRA 2026 and IEEE IV 2026, with ongoing work under review at NeurIPS 2026.

Sirjan University of Technology (2021–Present)
M.Sc. research in AI & Robotics, ranked 1st in cohort (GPA 19.9/20).
Work led to multiple first-author publications in top-tier venues.

Teaching

• Lecturer: Python Programming, English for Computer Science
• Designed ML Bootcamp (based on Andrew Ng ML Specialization)
• Mentored students in ML and research projects
• Organized Kaggle-style competitions and robotics challenges

Academic Service

• Peer Reviewer: IEEE Intelligent Vehicles Symposium (IV 2026)
• Workshop Speaker: Transformers in 3D Point Clouds (ICIS 2024)
• AI Scientific Student Association — Secretary

Contact

📧 Email
🎓 Google Scholar
💻 GitHub