Inside The wall

R0:97532435b0393f0a6ae72973cc68382e-How to avoid machine learning pitfalls: a guide for academic researchers

How to avoid machine learning pitfalls: a guide for academic researchers

This document gives a concise outline of some of the common mistakes that occur when using machine learning techniques, and what can be done to avoid them. It is intended primarily as a guide for research students, and focuses on issues that are of particular concern within academic research, such as the need to do rigorous comparisons and reach valid conclusions. It covers five stages of the machine learning process: what to do before model building, how to reliably build models, how to robustly evaluate models, how to compare models fairly, and how to report results

How to avoid machine learning pitfalls: a guide for academic researchers Leer más »

Páginas: 1 2 3
R0:94ecd1a3b2168ef2dbb43222545d08bc-YOLOX: Exceeding YOLO Series in 2021

YOLOX: Exceeding YOLO Series in 2021

We switch the YOLO detector to an anchor-free manner and conduct other advanced detection techniques, i.e., a decoupled head and the leading label assignment strategy SimOTA to achieve state-of-the-art results across a large scale range of models: For YOLO-Nano with only 0.91M parameters and 1.08G FLOPs, we get 25.3% AP on COCO, surpassing NanoDet by 1.8% AP; for YOLOv3, one of the most widely used detectors in industry, we boost it to 47.3% AP on COCO, outperforming the current best practice by 3.0% AP; for YOLOX-L with roughly the same amount of parameters as YOLOv4-CSP, YOLOv5-L, we achieve 50.0% AP on COCO at a speed of 68.9 FPS on Tesla V100, exceeding YOLOv5-L by 1.8% AP.

YOLOX: Exceeding YOLO Series in 2021 Leer más »

Páginas: 1 2 3
R0:b67b54ef535ceeaf4b3bf38c0cdf8c0b-Framework based on parameterized images on ResNet to identify intrusions in smartwatches or other related devices

Framework based on parameterized images on ResNet to identify intrusions in smartwatches or other related devices

The continuous appearance and improvement of mobile devices in the form of smartwatches, smartphones and other similar devices has led to a growing and unfair interest in putting their users under the magnifying glass and control of applications.

Framework based on parameterized images on ResNet to identify intrusions in smartwatches or other related devices Leer más »

Páginas: 1 2 3
R0:1bd6c89454b8a29952356211c7075d12-YOLOP: You Only Look Once for Panoptic Driving Perception

YOLOP: You Only Look Once for Panoptic Driving Perception

A panoptic driving perception system is an essential part of autonomous driving. A high-precision and real-time perception system can assist the vehicle in making the reasonable decision while driving. We present a panoptic driving perception network (YOLOP) to perform traffic object detection, drivable area segmentation and lane detection simultaneously. It is composed of one encoder for feature extraction and three decoders to handle the specific tasks. Our model performs extremely well on the challenging BDD100K dataset, achieving state-of-the-art on all three tasks in terms of accuracy and speed. Besides, we verify the effectiveness of our multi-task learning model for joint training via ablative studies.

YOLOP: You Only Look Once for Panoptic Driving Perception Leer más »

Páginas: 1 2 3
R0:2860421906bee85cbfa5eadd287f7a8c-Data as the main focus of “State of the art of data science in Spanish language and its application in the field of Artificial Intelligence”

Data as the main focus of “State of the art of data science in Spanish language and its application in the field of Artificial Intelligence”

According to the results, there is an evidence of cultural bias for data science in Spanish language. The outcome of the consultation, which carried out on 12 April 2021, confirms that only 10 out of 23.771 datasets “speaks” Spanish.»

Data as the main focus of “State of the art of data science in Spanish language and its application in the field of Artificial Intelligence” Leer más »

Páginas: 1 2 3
R0:a4d69dca3fc7b4bb1e9ce3992edead0d-'Framework' basado en imágenes parametrizadas sobre ResNet para identificar intrusiones en 'smartwatches' u otros dispositivos afines

‘Framework’ basado en imágenes parametrizadas sobre ResNet para identificar intrusiones en ‘smartwatches’ u otros dispositivos afines

La continua aparición y mejora de dispositivos móviles en forma de ‘smartwatches’, ‘smartphones’ y otros dispositivos similares ha propicio un creciente y desleal interés en poner bajo la lupa y el control de los aplicativos a sus usuarios. De forma ofuscada por los fabricantes.

‘Framework’ basado en imágenes parametrizadas sobre ResNet para identificar intrusiones en ‘smartwatches’ u otros dispositivos afines Leer más »

Páginas: 1 2 3
R0:379550808a93f299a00d8391ccd2e65c-Los datos como eje principal en el

Los datos como eje principal en el «Estado del arte de la ciencia de datos en el idioma español y su aplicación en el campo de la Inteligencia Artificial»

Los resultados de este estudio son una evidencia del sesgo cultural que existe entre la lengua inglesa y la española en la ciencia de datos. De los 23.771 conjuntos de datos que se encontraron con fecha de consulta 12/04/2021, tan solo 10 se encontraban en castellano

Los datos como eje principal en el «Estado del arte de la ciencia de datos en el idioma español y su aplicación en el campo de la Inteligencia Artificial» Leer más »

Páginas: 1 2 3