Deep Regression Models for Local Interaction in Multi-agent Robot Tasks

Sarmiento, Fredy Hernán Martínez; Penagos, Cristian; Pacheco, Luis

doi:10.1007/978-3-319-93818-9_7

Cited by 3 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The selection was made according to their high performance on similar problems, and the small size of the models that make them suitable for implementation on the robot. The research group had previously used these architectures in other modules of our robotic platform to solve problems such as automatic emotion recognition and movement strategies in unknown and dynamic environments [19], [20]. The selected architectures are ResNet, DenseNet, and NASNet.…”

Section: Methodsmentioning

confidence: 99%

“…Finally, it is possible to autonomously identify acoustic events by a small robot using convolutional neural networks (CNNs) because convolutional models can run in real-time after they have been trained correctly [19], [20]. Service robots must operate in human environments, interacting with humans, so they must be able to respond in real-time to immediate needs [21]- [23].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Acoustic event characterization for service robot using convolutional networks

Martinez

Sarmiento

Hernández

2022

IJECE

View full text Add to dashboard Cite

<p>This paper presents and discusses the creation of a sound event classification model using deep learning. In the design of service robots, it is necessary to include routines that improve the response of both the robot and the human being throughout the interaction. These types of tasks are critical when the robot is taking care of children, the elderly, or people in vulnerable situations. Certain dangerous situations are difficult to identify and assess by an autonomous system, and yet, the life of the users may depend on these robots. Acoustic signals correspond to events that can be detected at a great distance, are usually present in risky situations, and can be continuously sensed without incurring privacy risks. For the creation of the model, a customized database is structured with seven categories that allow to categorize a problem, and eventually allow the robot to provide the necessary help. These audio signals are processed to produce graphical representations consistent with human acoustic identification. These images are then used to train three convolutional models identified as high-performing in this type of problem. The three models are evaluated with specific metrics to identify the best-performing model. Finally, the results of this evaluation are discussed and analyzed.</p>

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Acoustic event characterization for service robot using convolutional networks

Martinez

Sarmiento

Hernández

2022

IJECE

View full text Add to dashboard Cite

show abstract

“…Deep networks have turned around the performance of systems in tasks such as speech recognition, computer vision, and language processing, applications in which traditional filtering-based strategies perform poorly (Caley et al, 2019, Martínez et al, 2018a. These learning techniques, in conjunction with reinforcement learning strategies, are crucial in pattern recognition applications (Hock & Schoelling, 2019, Martínez & Rendón, 2022.…”

Section: Introductionmentioning

confidence: 99%

A Deep-Fracture Approach to Embedded Multilayer Neural Networks for Handwritten Character Recognition

Sarmiento

2022

etj

Self Cite

View full text Add to dashboard Cite

Low-cost microcontrollers, but also with reduced processing and memory capabilities, are highly appreciated devices for the development of specific tasks as embedded systems. The devices found in the market today provide dedicated 32-bit architectures with communication capabilities ideal for IoT applications. Some of these tasks would highly benefit from the generalization and approximation capabilities of a neural network, but current convolutional networks prove to be too complex for microcontrollers. An alternative for some cases is the use of Multilayer Neural Networks (MNNs), which, thanks to a shallower depth, reach sizes suitable for use in small embedded systems. An MNN can accommodate a complex decision surface capable of being used in high dimensionality pattern classification problems, such as handwritten character recognition. Given the lack of an explicit strategy to define the architecture of these networks in coherence with the problem to be solved, this paper conducts a performance analysis of these networks in terms of their depth, to establish criteria to determine their size in a specific application. For the evaluation, the public database MNIST has been used in conjunction with classical metrics to evaluate the performance of a classification model. The results show that the depth of the network determines to a high degree the performance of the model, and guides the appropriate selection of the architecture.

show abstract

“…A widely used strategy both for the case of digital image processing (which is where it is, in fact, most used) and for the case of detection of specific sensors is the segmentation of the signal's field information (hand), and subsequent extraction of characteristics from each segment for subsequent classification [11,18]. One of the most widely used tools for the classification process is the neural networks, and more recently, the deep learning [19,20]. These tools have the advantage of being able to generalize specific characteristics of the signs and detect them in different hands and people, in a similar way as the human brain does [21].…”

mentioning

confidence: 99%

A gesture recognition system for the Colombian sign language based on convolutional neural networks

2020

View full text Add to dashboard Cite

Sign languages (or signed languages) are languages that use visual techniques, primarily with the hands, to transmit information and enable communication with deaf-mutes people. This language is traditionally only learned by people with this limitation, which is why communication between deaf and non-deaf people is difficult. To solve this problem we propose an autonomous model based on convolutional networks to translate the Colombian Sign Language (CSL) into normal Spanish text. The scheme uses characteristic images of each static sign of the language within a base of 24000 images (1000 images per category, with 24 categories) to train a deep convolutional network of the NASNet type (Neural Architecture Search Network). The images in each category were taken from different people with positional variations to cover any angle of view. The performance evaluation showed that the system is capable of recognizing all 24 signs used with an 88% recognition rate.

show abstract

Deep Regression Models for Local Interaction in Multi-agent Robot Tasks

Cited by 3 publications

References 9 publications

Acoustic event characterization for service robot using convolutional networks

Acoustic event characterization for service robot using convolutional networks

A Deep-Fracture Approach to Embedded Multilayer Neural Networks for Handwritten Character Recognition

A gesture recognition system for the Colombian sign language based on convolutional neural networks

Contact Info

Product

Resources

About