The reporter learned from Southwest University that the school’s research team has released a low-rank tensor recovery model, theory and algorithm based on binary quantization, which helps to improve the speed and accuracy of large-scale data transmission, while reducing hardware costs and enabling data transmission. Compression and preservation are more “cost-effective”. Related research results have been published online by the international journal “IEEE Model Analysis and Machine Intelligence Transactions” in the field of artificial intelligence.

According to the corresponding author of the paper and Professor Wang Jianjun from the School of Mathematics and Statistics of Southwest University, with the continuous expansion of the application of artificial intelligence technology, the amount of data transmission in the fields of image and video processing, pattern recognition, and computer vision is huge. However, due to hardware costs and requirements for data transmission speed, the currently commonly used low-rank tensor recovery (LRTR) method cannot achieve high-precision quantization of signals during large-scale data transmission. The resulting quantization error has a significant impact on system recovery performance. To influence. In response to this problem, Wang Jianjun’s research team proposed a low-rank tensor restoration model, theory and algorithm based on binary quantization. The principle is to combine the binary measurement method with low-rank tensor restoration so that the quantization process is incorporated into the model for processing , So as to effectively control the influence of quantization error on system recovery performance and make up for the defects of existing algorithms.

The test results show that this new algorithm has achieved higher restoration accuracy in practical applications such as face image restoration and multispectral image restoration, and the obtained image data is clearer. While the new algorithm reduces the hardware cost of data transmission, the data processing speed is also improved, making it possible to transmit, compress and save data with higher “cost-effectiveness”.

At present, Wang Jianjun’s research team has cooperated with related companies in the fields of mobile communication terminals and medical image processing to promote domestic mobile terminal technology updates, improve the processing speed of MRI in medical treatment, and reduce economic costs. The new algorithm is also expected to be combined with radar imaging technology to play a role in meteorological monitoring, geological prospecting and other fields.

Darlox Electronic Limited is mainly engaged in the design and manufacture of custom cables for flat-panel displays and the peripherals; custom cables for auto, mobile and office equipment; and also cables for almost all modern electronic products you can see. We cater to our customer’s every need with our design services and quick lead time. Orders of large quantity and small quantity are both always welcomed. In the era of IoT(Internet of Things), Darlox offers flex flat cables, flex PCB, micro coax cables assembly for data high speed transfer applications.