A5017568539- Advancements in PLL and VCO Technologies
- Advanced Neural Network Applications
- CCD and CMOS Imaging Sensors
- Advanced Memory and Neural Computing
- Advanced Electrical Measurement Techniques
- Advanced Frequency and Time Standards
- Brain Tumor Detection and Classification
- Advanced Optical Sensing Technologies
- Atomic and Subatomic Physics Research
- Photonic and Optical Devices
- Network Time Synchronization Technologies
- Image and Signal Denoising Methods
- Semiconductor Lasers and Optical Devices
- Anomaly Detection Techniques and Applications
- Sensor Technology and Measurement Systems
- Embedded Systems Design Techniques
- Advanced Image and Video Retrieval Techniques
- Neural Networks and Applications
- Radiomics and Machine Learning in Medical Imaging
- COVID-19 diagnosis using AI
Benemérita Universidad Autónoma de Puebla
2022-2024
Technological Institute of Celaya
2019-2020
Tecnológico Nacional de México
2020
This paper introduces a Trimmed-Tapped-Delay-Line-based Time-to-Digital Converter (TTDL-TDC) architecture. Unlike state-of-the-art works, where the propagation time of TDL is roughly equivalent to system clock period, proposed architecture can be shortened use only 63% necessary length propagate cycle without loss resolution. To achieve this reduction, signal propagated throughout Trimmed (TTDL), capturing its position with rising edge latching signal. An encoder based on transition...
Convolutional neural networks (CNN) have turned into one of the key algorithms in machine learning for content classification digital images. Nevertheless, CNN computational complexity is considerable larger than classic algorithms, thus, CPU- or GPU-based platforms are generally used implementations many applications, but often do not fulfill portable requirements due to resources, energy and real-time constrains. Therefore, there a growing interest on real time processing solutions object...
Convolutional Neuronal Networks (CNN) implementation on embedded devices is restricted due to the number of layers some CNN models. In this context, paper describes a novel architecture based Layer Operation Chaining (LOC) which uses fewer convolvers than convolution layers. A reutilization hardware promoted through kernel decomposition. Thus, an architectural design with reduced resources utilization achieved, suitable be implemented low-end as solution for portable classification...
Time-to-Digital Converters (TDCs) are instruments used to measure time intervals from digital signals accurately. They play an essential role in areas such as Light Detection and Ranging (LiDAR), Positron Emission To-mography (PET) [1] Particle Physics Experiments (PPE) [2]. Particularly, PPE involves the identification of particles resulting collisions accelerators. The estimation is carried out by measuring their Time Flight (ToF) deposited charge scintillator materials, becoming a...
A flexible systolic architecture to achieve hardware acceleration in the convolution operation for convolutional neural networks (CNNs) is described this paper. The main feature capability be adapted perform from different hyperparameters typically required CNN stages, like input map and filter sizes. proposed applied processing stages pretrained CNNs classify images size of 28x28 pixels MNIST database with sizes 3x3, 5x5, 7x7, 9x9 11x11. goal reduce data time without depending on a single...
This paper introduces a flexible convolver capable of adapting to the different convolution layer configurations state-of-the-art Convolution Neural Networks (CNNs). The use two proposed programmable components achieves this adaptability. A Programmable Line Buffer (PLB) based on Shift Registers (PSRs) allows generation required masks for each processed CNN layer. computing is performed through systolic array configured according target device resources. In order maximize resource usage and...