Compact data structures can represent data with usually a much smaller memory footprint than its plain representation. In addition to maintaining the data in a form that uses less space, they allow us to efficiently a...
详细信息
In the complex remote sensing image detection, there are still many challenges in maritime ship detection. This paper combines the latest swin-transformer model with satellite remote sensing big data, uses the SSDD da...
详细信息
A network of physical things that are outfitted with sensors, smart networking, and RFID technology is referred to as the "Internet of Things"(IoT), which is an abbreviation for the phrase. The items in ques...
详细信息
Millions of people across the globe were affected by this rapidly spreading disease by the year 2020. Contamination of the respiratory system results from contracting COVID-19. Nonetheless, establishing a conclusive d...
详细信息
In this paper, an acceleration measure (AM) for heart rate variability (HRV) is presented. Via the quantitative index AM, the acceleration of change of the normal sinus rhythm can be described effectively. Experiment ...
详细信息
Sickle cell disease (SCD) is a hereditary blood disorder that affects certain parts of the world. This disease affects hemoglobin, causing red blood cells to change shape, such as sickle and crescent, making it diffic...
详细信息
data augmentation technology has been widely used in computer vision and speech with good results. In computer vision and speech, simple manipulation of gold data can achieve great data augmentation effects. However, ...
详细信息
ISBN:
(纸本)9789819916474;9789819916481
data augmentation technology has been widely used in computer vision and speech with good results. In computer vision and speech, simple manipulation of gold data can achieve great data augmentation effects. However, in NLP (natural language processing), simple operations on data, such as "randomly delete words", "swap word positions", can have a huge impact on the semantics of sentences. This impact on semantics can be devastating for fine-grained tasks like NER (named entity recognition). In this work, we propose a novel model to generate diverse and high-quality data for NER, which is called DHQDA (Diverse and High-Quality data Augmentation). Our model outputs the data by using a small-scale neural network to prompt the key and value in the transformer block of the PLM (pre-trained language model). Our experimental results demonstrate that DHQDA performs more stable and better than baseline methods on both Chinese and English datasets, whether in rich or low-resource situations.
As many IoT devices generate an enormous and varied amount of data that need to be processed in a very short time, storing and processing IoT big data become a huge challenge. While lossy compression can dramatically ...
详细信息
ISBN:
(纸本)9791188428090
As many IoT devices generate an enormous and varied amount of data that need to be processed in a very short time, storing and processing IoT big data become a huge challenge. While lossy compression can dramatically reduce data volume, finding an optimal balance between volume reduction and information loss is not an easy task. The compression ratio is within a range tolerable by the application is crucial. Motivated by this, we analyze the characteristics of data compressed and present a prediction model about the compression ratio of transformation-based lossy compression algorithms for IoT datasets collected.
The introduction of an automation system in facilities has enabled the continuous flow of data streams, facilitating easier and more accessible data collection. Moreover, advancements in machine learning and deep lear...
详细信息
Vision-based object recognition is an important enabler for automating specific workflows in production and transportation scenarios. Locating and manipulating pallets as common functional objects represents a relevan...
详细信息
暂无评论