No abstract

Video super-resolution is one of the most popular tasks on mobile devices, being widely used for an automatic improvement of lowbitrate and low-resolution video streams. While numerous solutions have been proposed for this problem, they are usually quite computationally demanding, demonstrating low FPS rates and power efficiency on mobile devices. In this Mobile AI challenge, we address this problem and propose the participants to design an end-to-end real-time video super-resolution solution for mobile NPUs optimized for low energy consumption. The participants were provided with the REDS training dataset containing video sequences for a 4X video upscaling task. The runtime and power efficiency of all models was evaluated on the powerful MediaTek Dimensity 9000 platform with a dedicated AI processing unit capable of accelerating floating-point and quantized neural networks. All proposed solutions are fully compatible with the above NPU, demonstrating an up to 500 FPS rate and 0.2 [Watt / 30 FPS] power consumption. A detailed description of all models developed in the challenge is provided in this paper.

This paper introduces the real image Super-Resolution (SR) challenge that was part of the Advances in Image Manipulation (AIM) workshop, held in conjunction with ECCV 2020. This challenge involves three tracks to super-resolve an input image for ×2, ×3 and ×4 scaling factors, respectively. The goal is to attract more attention to realistic image degradation for the SR task, which is much more complicated and challenging, and contributes to real-world image super-resolution applications. 452 participants were registered for three tracks in total, and 24 teams submitted their results. They gauge the state-of-the-art approaches for real image SR in terms of PSNR and SSIM.

Abstract-Traffic estimation is an important issue to analyze the traffic congestion in large-scale urban traffic situations.Recently, many researchers have used GPS data to estimate traffic congestion. However, how to fuse the multiple data reasonably and guarantee the accuracy and efficiency of these methods are still challenging problems. In this paper, we propose a novel method Multiple Data Estimation (MDE) to estimate the congestion status in urban environment with GPS trajectory data efficiently, where we estimate the congestion status of the area through utilizing multiple properties, including density, velocity, inflow and previous status. Among them, traffic inflow and previous status (combination of time and space factors) are not both used in other existing methods. In order to ensure the accuracy and efficiency, we apply dynamic weights of data and parameters in MDE method. To evaluate our methods, we apply it on large-scale taxi GPS data of Beijing and Shanghai. Extensive experiments on these two real-world datasets demonstrate the significant improvements of our method over several state-ofthe-art methods.

The crow search algorithm (CSA) is a new intelligent optimization algorithm based on the behavior of the crow population, which has been proven to perform well. However, its simple search mechanism also leads to the algorithm's slow convergence speed and its ease of falling into local optimization when solving complex optimization problems. In order to overcome these problems, this paper proposes an improved CSA (ISCSA) based on a spiral search mechanism. By introducing a weight coefficient, an optimal guidance position and a spiral search mechanism, the position equation was updated to accelerate the convergence of the algorithm and make the exploration and exploitation of CSA more balanced. Meanwhile, adding Gaussian variation and random perturbation strategy made it difficult for the algorithm to fall into local optimization. The advantages of the proposed ISCSA were evaluated using 23 benchmark functions and four classical engineering design problems. The experimental and statistical results of 23 test functions showed that the proposed ISCSA could escape from the local optima with higher accuracy and faster convergence than both the CSA and other meta-heuristic optimization algorithms. The calculation results of the four engineering optimization problems showed that compared with other algorithms, ISCSA can solve the practical optimization problem well and has been proved to have strong competitiveness and good performance. INDEX TERMS Crow search algorithm, weight coefficient, spiral search mechanism, engineering design optimization.

Recognition of human intention based on Electroencephalography (EEG) signals attracts strong research interest in pattern recognition because of its promising applications that enable non-muscular communications and controls. Over the past few years, most EEG-based recognition works make significant efforts to learn extracted features to explore specific patterns between a segment of EEG signals and the corresponding activities. Unfortunately, vectorization-based feature representations, either vector-like or matrix-like ones, suffer from massive signal noise and difficulties of exploiting signal correlations between adjacent sensors of EEG signals. Most importantly, EEG signals are represented by one unique frequency and then fed into the subsequent learning model. Neglecting different frequencies of EEG signals can be detrimental to activity recognition because a particular frequency of EEG signals is more helpful to recognize some activities. Inspired by this idea, we propose to extract EEG signals with different frequencies and introduce a novel Multi-task deep learning model to learn the human intentions. We have conducted extensive experiments on a publicly available EEG benchmark dataset and compared our method with many state-of-the-art algorithms. The experimental results demonstrate that the proposed Multi-task deep recurrent neural network outperforms all the compared methods in a multi-class scenario.

No abstract

No abstract