LSSVC: A Learned Spatially Scalable Video Coding Scheme.

Traditional block-based spatially scalable video coding has been studied for over twenty years. While significant advancements have been made, the scope for further improvement in compression performance is limited. Inspired by the success of learned video coding, we propose an end-to-end learned spatially scalable video coding scheme, LSSVC, which provides a new solution for scalable video coding. In LSSVC, we propose to use the motion, texture, and latent information of the base layer (BL) as interlayer information for compressing the enhancement layer (EL). To reduce interlayer redundancy, we design three modules to leverage the upsampled interlayer information. Firstly, we design a contextual motion vector (MV) encoder-decoder, which utilizes the upsampled BL motion information to help compress high-resolution MV. Secondly, we design a hybrid temporal-layer context mining module to learn more accurate contexts from the EL temporal features and the upsampled BL texture information. Thirdly, we use the upsampled BL latent information as an interlayer prior for the entropy model to estimate more accurate probability distribution parameters for the high-resolution latents. Experimental results show that our scheme surpasses H.265/SHVC reference software by a large margin. Our code is available at https://github.com/EsakaK/LSSVC.

Estimation of Weight and Body Measurement Model for Pigs Based on Back Point Cloud Data.

Pig farming is a crucial sector in global animal husbandry. The weight and body dimension data of pigs reflect their growth and development status, serving as vital metrics for assessing their progress. Presently, pig weight and body dimensions are predominantly measured manually, which poses challenges such as difficulties in herding, stress responses in pigs, and the control of zoonotic diseases. To address these issues, this study proposes a non-contact weight estimation and body measurement model based on point cloud data from pig backs. A depth camera was installed above a weighbridge to acquire 3D point cloud data from 258 Yorkshire-Landrace crossbred sows. We selected 200 Yorkshire-Landrace sows as the research subjects and applied point cloud filtering and denoising techniques to their three-dimensional point cloud data. Subsequently, a K-means clustering segmentation algorithm was employed to extract the point cloud corresponding to the pigs' backs. A convolutional neural network with a multi-head attention was established for pig weight prediction and added RGB information as an additional feature. During the data processing process, we also measured the back body size information of the pigs. During the model evaluation, 58 Yorkshire-Landrace sows were specifically selected for experimental assessment. Compared to manual measurements, the weight estimation exhibited an average absolute error of 11.552 kg, average relative error of 4.812%, and root mean square error of 11.181 kg. Specifically, for the MACNN, incorporating RGB information as an additional feature resulted in a decrease of 2.469 kg in the RMSE, a decrease of 0.8% in the MAPE, and a decrease of 1.032 kg in the MAE. Measurements of shoulder width, abdominal width, and hip width yielded corresponding average relative errors of 3.144%, 3.798%, and 3.820%. In conclusion, a convolutional neural network with a multi-head attention was established for pig weight prediction, and incorporating RGB information as an additional feature method demonstrated accuracy and reliability for weight estimation and body dimension measurement.

Understanding Naturalistic Facial Expressions with Deep Learning and Multimodal Large Language Models.

This paper provides a comprehensive overview of affective computing systems for facial expression recognition (FER) research in naturalistic contexts. The first section presents an updated account of user-friendly FER toolboxes incorporating state-of-the-art deep learning models and elaborates on their neural architectures, datasets, and performances across domains. These sophisticated FER toolboxes can robustly address a variety of challenges encountered in the wild such as variations in illumination and head pose, which may otherwise impact recognition accuracy. The second section of this paper discusses multimodal large language models (MLLMs) and their potential applications in affective science. MLLMs exhibit human-level capabilities for FER and enable the quantification of various contextual variables to provide context-aware emotion inferences. These advancements have the potential to revolutionize current methodological approaches for studying the contextual influences on emotions, leading to the development of contextualized emotion models.

A green and effective room-temperature recycling process of LiFePO4 cathode materials for lithium-ion batteries.

Nowadays LiFePO4 cathode develops rapidly for its advantages of long life span, low cost and non-toxicity, especially in electrical vehicle markets. Because of its stable olivine structure, LiFePO4 is difficult to be recycled by the conventional hydrometallurgical processes as for LiCoO2 or LiNixCoyMnzO2. Pyrometallurgical processes consume much energy and release toxic gases. Herein, an effective room-temperature process based on the mechanochemical treatment is proposed to extract metals from LiFePO4. Spent LiFePO4 is co-grinded with low-cost citric acid agent in a ball mill. After grinding, the mixture is dissolved in deionized water and filtrated. With addition of H2O2, the extraction efficiency of Li reaches as high as 99.35%. Conversely, Fe is hardly extracted with a low extraction efficiency of only 3.86%, indicating a selective recovery of valuable Li element. In addition, when H2O is used instead of H2O2, the mechanochemical reaction changes and the extraction efficiencies of Li and Fe at optimal conditions reach 97.82% and 95.62%, respectively. The Fe impurity is removed as Fe(OH)3 precipitation by adding NaOH, and Li is recycled as Li2CO3 after reaction with saturated Na2CO3 at 95 °C. This simple and easily-operated process has little negative impact on the environment and has great potential in industrial applications.

Toward sustainable and systematic recycling of spent rechargeable batteries.

Ever-growing global energy needs and environmental damage have motivated the pursuit of sustainable energy sources and storage technologies. As attractive energy storage technologies to integrate renewable resources and electric transportation, rechargeable batteries, including lead-acid, nickel-metal hydride, nickel-cadmium, and lithium-ion batteries, are undergoing unprecedented rapid development. However, the intrinsic toxicity of rechargeable batteries arising from their use of toxic materials is potentially environmentally hazardous. Additionally, the massive production of batteries consumes numerous resources, some of which are scarce. It is therefore essential to consider battery recycling when developing battery systems. Here, we provide a systematic overview of rechargeable battery recycling from a sustainable perspective. We present state-of-the-art fundamental research and industrial technologies related to battery recycling, with a special focus on lithium-ion battery recycling. We introduce the concept of sustainability through a discussion of the life-cycle assessment of battery recycling. Considering the forecasted trend of a massive number of retired power batteries from the forecasted surge in electric vehicles, their repurposing and reuse are considered from economic, technical, environmental, and market perspectives. New opportunities, challenges, and future prospects for battery recycling are then summarized. A reinterpreted 3R strategy entailing redesign, reuse, and recycling is recommended for the future development of battery recycling.

Process for recycling mixed-cathode materials from spent lithium-ion batteries and kinetics of leaching.

A "grave-to-cradle" process for the recycling of spent mixed-cathode materials (LiCoO2, LiCo1/3Ni1/3Mn1/3O2, and LiMn2O4) has been proposed. The process comprises an acid leaching followed by the resynthesis of a cathode material from the resulting leachate. Spent cathode materials were leached in citric acid (C6H8O7) and hydrogen peroxide (H2O2). Optimal leaching conditions were obtained at a leaching temperature of 90 °C, a H2O2 concentration of 1.5 vol%, a leaching time of 60 min, a pulp density of 20 g L-1, and a citric acid concentration of 0.5 M. The leaching efficiencies of Li, Co, Ni, and Mn exceeded 95%. The leachate was used to resynthesize new LiCo1/3Ni1/3Mn1/3O2 material by using a sol-gel method. A comparison of the electrochemical properties of the resynthesized material (NCM-spent) with that synthesized directly from original chemicals (NCM-syn) indicated that the initial discharge capacity of NCM-spent at 0.2 C was 152.8 mA h g-1, which was higher than the 149.8 mA h g-1 of NCM-syn. After 160 cycles, the discharge capacities of the NCM-spent and NCM-syn were 140.7 mA h g-1 and 121.2 mA h g-1, respectively. After discharge at 1 C for 300 cycles, the NCM-spent material remained a higher capacity of 113.2 mA h g-1 than the NCM-syn (78.4 mA h g-1). The better performance of the NCM-spent resulted from trace Al doping. A new formulation based on the shrinking-core model was proposed to explain the kinetics of the leaching process. The activation energies of the Li, Co, Ni, and Mn leaching were calculated to be 66.86, 86.57, 49.46, and 45.23 kJ mol-1, respectively, which indicates that the leaching was a chemical reaction-controlled process.