Classification guided thick fog removal network for drone imaging: ClassifyCycle.

The foggy images captured by drones are nonuniform due to inhomogeneous distribution of fog in higher altitude, leading to the obvious fog thickness differences in the images. This paper proposes a classification guided thick fog removal network for drone imaging, termed ClassifyCycle. The drone images are input into the proposed classification module (ICLFn) to enhance the reliability of follow-up learning network. The style migration module (ISMn) is introduced to reduce the image distortion, such as hue artifact and texture distort. The proposed network ClassifyCycle does not require paired foggy and corresponding fog-free datasets, avoiding the phenomena of overexposure, distortion, color deviation and fog residue after defogging. Extensive experimental results show that the proposed ClassifyCycle network surpasses the state-of-the-art algorithms on synthetic and realistic drone images captured in thick fog weather.

E-YOLOv4-tiny: a traffic sign detection algorithm for urban road scenarios.

Introduction: In urban road scenes, due to the small size of traffic signs and the large amount of surrounding interference information, current methods are difficult to achieve good detection results in the field of unmanned driving. Methods: To address the aforementioned challenges, this paper proposes an improved E-YOLOv4-tiny based on the YOLOv4-tiny. Firstly, this article constructs an efficient layer aggregation lightweight block with deep separable convolutions to enhance the feature extraction ability of the backbone. Secondly, this paper presents a feature fusion refinement module aimed at fully integrating multi-scale features. Moreover, this module incorporates our proposed efficient coordinate attention for refining interference information during feature transfer. Finally, this article proposes an improved S-RFB to add contextual feature information to the network, further enhancing the accuracy of traffic sign detection. Results and discussion: The method in this paper is tested on the CCTSDB dataset and the Tsinghua-Tencent 100K dataset. The experimental results show that the proposed method outperforms the original YOLOv4-tiny in traffic sign detection with 3.76% and 7.37% improvement in mAP, respectively, and 21% reduction in the number of parameters. Compared with other advanced methods, the method proposed in this paper achieves a better balance between accuracy, real-time performance, and the number of model parameters, which has better application value.

State of Health Estimation Based on the Long Short-Term Memory Network Using Incremental Capacity and Transfer Learning.

Battery state of health (SOH) estimating is essential for the safety and preservation of electric vehicles. The degradation mechanism of batteries under different aging conditions has attracted considerable attention in SOH prediction. In this article, the discharge voltage curve early in the cycle is considered to be strongly characteristic during cell aging. Therefore, the battery aging state can be quantitatively characterized by an incremental capacity analysis (ICA) of the voltage distribution. Due to the interference of vibration noise of the test platform, the discrete wavelet transform (DWT) methods are accustomed to soften the premier incremental capacity curves in different hierarchical decompositions. By analyzing the battery aging mechanism, the peak of the curve and its corresponding voltage are used in the characterization of capacity decay by grey relation analysis (GRA) and to optimize the input of the deep learning model, and finally, the double-layer long short-term memory network (LSTM) model is used to train the data. The results demonstrate that the proposed model can predict the SOH of a single battery cycle using only small batch data and the relative error is less than 2%. Further, by freezing the LSTM layer for transfer learning, it can be used for battery health estimation in different loading modes. The results of training and verification show that this method has high accuracy and reliability in SOH estimation.

Cooperative Perception Technology of Autonomous Driving in the Internet of Vehicles Environment: A Review.

Cooperative perception, as a critical technology of intelligent connected vehicles, aims to use wireless communication technology to interact and fuse environmental information obtained by edge nodes with local perception information, which can improve vehicle perception accuracy, reduce latency, and eliminate perception blind spots. It has become a current research hotspot. Based on the analysis of the related literature on the Internet of vehicles (IoV), this paper summarizes the multi-sensor information fusion method, information sharing strategy, and communication technology of autonomous driving cooperative perception technology in the IoV environment. Firstly, cooperative perception information fusion methods, such as image fusion, point cloud fusion, and image-point cloud fusion, are summarized and compared according to the approaches of sensor information fusion. Secondly, recent research on communication technology and the sharing strategies of cooperative perception technology is summarized and analyzed in detail. Simultaneously, combined with the practical application of V2X, the influence of network communication performance on cooperative perception is analyzed, considering factors such as latency, packet loss rate, and channel congestion, and the existing research methods are discussed. Finally, based on the summary and analysis of the above studies, future research issues on cooperative perception are proposed, and the development trend of cooperative perception technology is forecast to help researchers in this field quickly understand the research status, hotspots, and prospects of cooperative perception technology.

Concurrent Topology Optimization for Maximizing the Modal Loss Factor of Plates with Constrained Layer Damping Treatment.

Damping performance of the plates with constrained layer damping (CLD) treatment mainly depends on the layout of CLD material and the material physical properties of the viscoelastic damping layer. This paper develops a concurrent topology optimization methodology for maximizing the modal loss factor (MLF) of plates with CLD treatment. At the macro scale, the damping layer is composed of 3D periodic unit cells (PUC) of cellular viscoelastic damping materials. At the micro scale, due to the deformation of viscoelastic damping material affected by the base and constrained layers, the representative volume element (RVE) considering a rigid skin effect is used to improve the accuracy of the effective constitutive matrix of the viscoelastic damping material. Maximizing the MLFs of CLD plates is employed as the design objectives in optimization procedure. The sensitivities with respect to macrodesign variables are formulated using the adjoint vector method while considering the contribution of eigenvectors, while the influence of macroeigenvectors is ignored to improve the computational efficiency in the mesosensitivity analysis. The macro and meso scales design variables are simultaneously updated using the Method of Moving Asymptotes (MMA) to find concurrently optimal configurations of constrained and viscoelastic damping layers at the macro scale and viscoelastic damping materials at the micro scale. Two rectangular plates with different boundary conditions are presented to validate the optimization procedure and demonstrate the effectiveness of the proposed concurrent topology optimization approach. The effects of optimization objectives and volume fractions on the design results are investigated. The results indicate that the optimized layouts of the macrostructure are dependent on the objective mode and the volume fraction on the meso scale. The optimized designs on the meso scale are mainly related to the objective mode. By varying the volume fraction on the macro scale, the optimized designs on the meso scale are different only in their detailed size, which is reflected in the values of the equivalent constitutive matrices.

Lower limb joint motion and muscle force in treadmill and over-ground exercise.

BACKGROUND: Treadmill exercise is commonly used as an alternative to over-ground walking or running. Increasing evidence indicated the kinetics of treadmill exercise is different from that of over-ground. Biomechanics of treadmill or over-ground exercises have been investigated in terms of energy consumption, ground reaction force, and surface EMG signals. These indexes cannot accurately characterize the musculoskeletal loading, which directly contributes to tissue injuries. This study aimed to quantify the differences of lower limb joint angles and muscle forces in treadmills and over-ground exercises. 10 healthy volunteers were required to walk at 100 and 120 steps/min and run at 140 and 160 steps/min on treadmill and ground. The joint flexion angles were obtained from the motion capture experiments and were used to calculate the muscle forces with an inverse dynamic method. RESULTS: Hip, knee, and ankle joint motions of treadmill and over-ground conditions were similar in walking, yet different in running. Compared with over-ground running, joint motion ranges in treadmill running were smaller. They were also less affected by stride frequency. Maximum Gastrocnemius force was greater in treadmill walking, yet maximum Rectus femoris and Vastus forces were smaller. Maximum Gastrocnemius and Soleus forces were greater in treadmill running. CONCLUSIONS: Treadmill exercise results in smoother joint kinematics. In terms of muscle force, treadmill exercise requires lower loading on knee extensor, yet higher loading on plantar flexor, especially on Gastrocnemius. The findings and the methodology can provide the basis for rehabilitation therapy customization and sophistic treadmill design.

Rich biotin content in lignocellulose biomass plays the key role in determining cellulosic glutamic acid accumulation by Corynebacterium glutamicum.

BACKGROUND: Lignocellulose is one of the most promising alternative feedstocks for glutamic acid production as commodity building block chemical, but the efforts by the dominant industrial fermentation strain Corynebacterium glutamicum failed for accumulating glutamic acid using lignocellulose feedstock. RESULTS: We identified the existence of surprisingly high biotin concentration in corn stover hydrolysate as the determining factor for the failure of glutamic acid accumulation by Corynebacterium glutamicum. Under excessive biotin content, induction by penicillin resulted in 41.7 ± 0.1 g/L of glutamic acid with the yield of 0.50 g glutamic acid/g glucose. Our further investigation revealed that corn stover contained 353 ± 16 µg of biotin per kg dry solids, approximately one order of magnitude greater than the biotin in corn grain. Most of the biotin remained stable during the biorefining chain and the rich biotin content in corn stover hydrolysate almost completely blocked the glutamic acid accumulation. This rich biotin existence was found to be a common phenomenon in the wide range of lignocellulose biomass and this may be the key reason why the previous studies failed in cellulosic glutamic acid fermentation from lignocellulose biomass. The extended recording of the complete members of all eight vitamin B compounds in lignocellulose biomass further reveals that the major vitamin B members were also under the high concentration levels even after harsh pretreatment. CONCLUSIONS: The high content of biotin in wide range of lignocellulose biomass feedstocks and the corresponding hydrolysates was discovered and it was found to be the key factor in determining the cellulosic glutamic acid accumulation. The highly reserved biotin and the high content of their other vitamin B compounds in biorefining process might act as the potential nutrients to biorefining fermentations. This study creates a new insight that lignocellulose biorefining not only generates inhibitors, but also keeps nutrients for cellulosic fermentations.

Tolerance improvement of Corynebacterium glutamicum on lignocellulose derived inhibitors by adaptive evolution.

Robustness of fermenting strains to lignocellulose derived inhibitors is critical for efficient biofuel and biochemical productions. In this study, the industrial fermenting strain Corynebacterium glutamicum S9114 was evolved for improved inhibitor tolerance using long-term adaptive evolution by continuously transferring into the inhibitors containing corn stover hydrolysate every 24 h, and finally a stably evolved C. glutamicum was obtained after 128 days of serial transfers. The evolved strain exhibited the highly increased conversion rate to the typical lignocellulose derived inhibitors including furfural, 5-hydroxymethylfurfural, vanillin, syringaldehyde, 4-hydroxybenzaldehyde, and acetic acid. Glucose consumption was obviously accelerated, and 22.4 g/L of glutamic acid was achieved in the corn stover hydrolysate, approximately 68.4% greater than that by the original strain. Whole genome re-sequencing revealed various mutations with the potential connection to the improved performance of the evolved strain. Transcriptional analysis further demonstrated that the glucose-PTS transport and the pentose phosphate pathway were significantly upregulated in the evolved strain, which very likely contributed to the accelerated glucose consumption, as well as sufficient NAD(P)H supply for aldehyde inhibitors reduction conversion and thus enhanced the inhibitor tolerance. This study provided important experimental evidences and valuable genetic information for robust strain construction and modification in lignocellulose biorefining processes.

Metformin prevents cerebellar granule neurons against glutamate-induced neurotoxicity.

Metformin, a wildly used drug for type 2 diabetes, has recently been proven to protect a variety of cells from stress including stroke. Glutamate is an excitatory neurotransmitter that contributes to excitatory neuronal damage involved in stroke and neurodegenerative disorders. In this study, we demonstrated that pretreatment of rat cerebellar granule neurons (CGN) with metformin greatly enhanced cell viability against glutamate-induced neurotoxicity. Metformin significantly attenuated neuronal apoptosis in glutamate-treated CGN by reducing cytochrome c releasing, caspase-3 activation and phosphorylation of MAP kinases. Our results suggested that metformin was able to directly inhibit glutamate induced excitotoxicity in neurons and might be beneficial to patients suffered from stroke and neurodegenerative disorders.