Development of a Highly Precise Place Recognition Module for Effective Human-robot Interactions in Changing Lighting and Viewpoint Conditions

Abstract

We present a highly precise and robust module for indoor place recognition, extending the work by Lemaignan et al. and Robert Jr. by giving the robot the ability to recognize its environment context. We developed a full end-to-end convolutional neural network architecture, using a pre-trained deep convolutional neural network and the explicit inductive bias transfer learning strategy. Experimental results based on the York University and Rzeszów University dataset show excellent performance values (over 94.75 and 97.95 percent accuracy) and a high level of robustness over changes in camera viewpoint and lighting conditions, outperforming current benchmarks. Furthermore, our architecture is 82.46 percent smaller than the current benchmark, making our module suitable for embedding into mobile robots and easily adoptable to other datasets without the need for heavy adjustments.

High-Performance Detection of Corneal Ulceration Using Image Classification with Convolutional Neural Networks

“…The combination of automated feature extraction and final classification represents one of the major advantages of CNNs [18,38]. Regarding the objective of reducing negative human factors such as bias, fatigue, and mindset during the medical assessment process to achieve objective and reproducible results [9,33], the automated feature extraction represents a fundamental backing for our approach.…”

“…Subsequently, we added our classifier to the base model, consisting of two dense layers with 64 and 32 units, each followed by a dropout layer with a rate of 0.5 [46,47]. We also used an appropriate L2 weight regularization with a weight decay of 1e-4 (alpha), ensuring the preservation of the prior network knowledge and preventing our model from overfitting [38].…”

High-performance Diagnosis of Sleep Disorders: A Novel, Accurate and Fast Machine Learning Approach Using Electroencephalographic Data

“…Despite we intensively evaluated other traditional machine learning approaches such as clustering [37] and also most modern convolutional neural networks, which are outstanding in other domains such as image Page 3251 recognition [38][39][40], we achieved the best results here with our novel method, originally proposed in [41,42]. However, future work should extend the application of further novel machine learning approaches.…”

Development of a Machine Learning Based Algorithm To Accurately Detect Schizophrenia based on One-minute EEG Recordings

“…While we intensively evaluated other traditional machine learning approaches such as clustering [46] and also most modern convolutional neural networks, which are outstanding in other domains such as image recognition [47][48][49], we achieved the best results here with our novel tree-based method proposed in [34]. However, the method of choice always limits scientific understanding.…”