Size Does Matter: Overcoming Limitations during Training when using a Feature Pyramid Network
Authors:
- 1 December 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1553-1560
Abstract
State-of-the-art object detectors need to be trained with a wide variety of data in order to perform well in real-world problems. Training-data-diversity is very important to achieve good generalization. However, there are scenarios where we have training data with certain limitations. One such scenario is when the objects of the testing set have a different size (discrepancy) from the objects used during training. Another scenario is when we have high-resolution images with a dimension that is not supported by the model. To address these problems, we propose a novel pipeline that is able to handle high-resolution images by cropping the original image into sub-images and put them back in the end. Also, in the case of the discrepancy of object sizes, we propose two different techniques based on scaling the image up and down in order to have an acceptable performance. In addition, we also use the information from the Feature Pyramid Network to remove false-positives. Our proposed methods overcome state-of-the-art data augmentation policies and our models can generalize to different object sizes even though limited data is provided.Keywords
This publication has 23 references indexed in Scilit:
- A New Approach to Compute CNNs for Extremely Large Images2017
- Aggregated Residual Transformations for Deep Neural Networks2017
- Negotiating Agriculture in the World Trade Organization: Food Security as a Non-Trade ConcernRevista de Direito Internacional, 2017
- Deep Count: Fruit Counting Based on Deep Simulated LearningSensors, 2017
- DeepFruits: A Fruit Detection System Using Deep Neural NetworksSensors, 2016
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- IMPORTANCE OF AGRICULTURE WITHIN THE STRUCTURE OF EMPLOYMENT AND PRODUCTION IN THE MEDITERRANEAN COUNTRIESJournal of Agribusiness and Rural Development, 2015
- Classification of Fruits Using Computer Vision and a Multiclass Support Vector MachineSensors, 2012
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009
- SOME INTERESTING SPECIAL CASES OF A NON-LOCAL PROBLEM MODELLING OHMIC HEATING WITH VARIABLE THERMAL CONDUCTIVITYProceedings of the Edinburgh Mathematical Society, 2001