Summary of Transfer Learning Approach For Railway Technical Map (rtm) Component Identification, by Obadage Rochana Rumalshan et al.
Transfer Learning Approach for Railway Technical Map (RTM) Component Identification
by Obadage Rochana Rumalshan, Pramuka Weerasinghe, Mohamed Shaheer, Prabhath Gunathilake, Erunika Dayaratna
First submitted to arxiv on: 21 May 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Digital Libraries (cs.DL)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed system uses Deep Learning and Optical Character Recognition techniques to digitize railway map component data from images and generate formatted text files. This research aims to address the inefficiency of current Railway Technical Maps (RTMs) available only in PDF format. The study compares three object detection models – YOLOv3, SSD, and Faster-RCNN – and finds that Faster-RCNN yields the highest mean Average Precision (mAP) and F1 score values. Furthermore, it is demonstrated that OCR performance can be improved by applying a pre-processing pipeline to remove distortions from text-containing images. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you’re trying to read a map on your phone, but it’s in a weird format and hard to understand. This paper solves this problem by using special computer programs (Deep Learning and Optical Character Recognition) to turn images of railway maps into easy-to-read text files. Right now, these maps are only available as PDFs, which can be tricky to work with. The researchers tested three different ways to recognize objects on the map (YOLOv3, SSD, and Faster-RCNN) and found that one method works better than the others. They also showed that making small changes to how they processed the images helped improve their results. |
Keywords
» Artificial intelligence » Deep learning » F1 score » Faster rcnn » Mean average precision » Object detection
