Indarto, Novelio Putra
Unknown Affiliation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Camera-based simultaneous localization and mapping: methods, camera types, and deep learning trends Dwimantara, Anak Agung Ngurah Bagus; Natan, Oskar; Indarto, Novelio Putra; Dharmawan, Andi
IAES International Journal of Robotics and Automation (IJRA) Vol 14, No 2: June 2025
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijra.v14i2.pp162-172

Abstract

The development of simultaneous localization and mapping (SLAM) technology is crucial for advancing autonomous systems in robotics and navigation. However, camera-based SLAM systems face significant challenges in accuracy, robustness, and computational efficiency, particularly under conditions of environmental variability, dynamic scenes, and hardware limitations. This paper provides a comprehensive review of camera-based SLAM methodologies, focusing on their different approaches for pose estimation, map reconstruction, and camera type. The application of deep learning also will be discussed on how it is expected to improve performance. The objective of this paper is to advance the understanding of camera-based SLAM systems and to provide a foundation for future innovations in robust, efficient, and adaptable SLAM solutions. Additionally, it offers pertinent references and insights for the design and implementation of next-generation SLAM systems across various applications.
Edge-aware distilled segmentation with pseudo-label refinement for autonomous driving perception Indarto, Novelio Putra; Natan, Oskar; Dharmawan, Andi
IAES International Journal of Robotics and Automation (IJRA) Vol 14, No 3: December 2025
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijra.v14i3.pp376-386

Abstract

Achieving precise semantic segmentation is essential for enabling real-time perception in autonomous systems, yet leading approaches typically require substantial annotated data and powerful hardware, restricting their use on devices with limited resources. This work introduces an efficient segmentation framework that integrates pseudo-label refinement, knowledge distillation, and entropy-based confidence filtering to train compact student networks suitable for edge deployment. High-quality pseudo-labels are first produced by a robust teacher network, then further improved using a dense conditional random field to boost spatial consistency. An entropy-based selection mechanism removes unreliable predictions, ensuring that only the most trustworthy labels guide the student model's training. The use of knowledge distillation effectively transfers detailed semantic understanding from the teacher to the student, enhancing accuracy without added computational overhead. Experimental results with multiple EfficientNet backbones reveal that this pipeline improves segmentation accuracy and output clarity, while also supporting real-time or near real-time inference on CPUs with limited processing power. Extensive ablation and qualitative studies further confirm the method's robustness and flexibility for real-world edge applications.