Abstract: The present disclosure relates to methods and apparatuses for managing a shared cache. One example method includes determining an access characteristic of accessing the shared cache by IO requests of each of K types that access the shared cache, determining a partition size and an eviction algorithm of the IO requests of each type in the shared cache based on the determined access characteristic and a hit rate of the shared cache, and configuring a cache size of the IO requests of each type in the shared cache as the determined partition size, and configuring an eviction algorithm of the IO requests of each type in the shared cache as the determined eviction algorithm.

Abstract: A data deduplication method includes: receiving a write request with first data block; writing the first data block into a storage device; writing metadata of the first data block into a first partition that is in a plurality of partitions of a metadata management structure and that is determined based on a feature of the first data block; and deleting the metadata of the first data block in the first partition and deleting the first data block from the storage device based on address information of the first data block when a fingerprint that is the same as the fingerprint of the first data block exists in the first partition. The method can prevent infrequently updated data from being evicted because resources are occupied by frequently updated data, thereby improving a deduplication ratio.

Abstract: A non-contact facial blood pressure measurement method based on 3D CNN is disclosed, which belongs to the technical field of computer vision. The method includes the following steps. S110: collecting an actual face video sample and training a blood pressure prediction model based on face images using 3D CNN neural network. S120: obtaining a face video in real time through a HD camera. S130: recognizing face key points in the face video obtained in S120 through dlib face recognition model, selecting a face region of interest, and extracting face images from the region. S140: performing a wavelet transform operation on the face images extracted in S130 to remove noise. S150: inputting seven consecutive frames of the face images into the 3D CNN blood pressure prediction model trained in S110 to obtain a blood pressure value of the measured person. The disclosure realizes non-contact facial blood pressure measurement.

Abstract: The present disclosure relates to methods and apparatuses for managing a shared cache. One example method includes determining an access characteristic of accessing the shared cache by IO requests of each of K types that access the shared cache, determining a partition size and an eviction algorithm of the IO requests of each type in the shared cache based on the determined access characteristic and a hit rate of the shared cache, and configuring a cache size of the IO requests of each type in the shared cache as the determined partition size, and configuring an eviction algorithm of the IO requests of each type in the shared cache as the determined eviction algorithm.

Abstract: The present disclosure belongs to the technical field of liquid/liquid interface electrochemistry and analytical chemistry, and specifically provides construction of a microtube-based ionic liquid colloid/water interface and use. In the present disclosure, the construction method of an ionic liquid colloid/water interface with a high stability and a desirable selectivity includes the following steps: adding a poly(ionic liquid) into an ionic liquid to form an ionic liquid colloid, to enhance an interfacial stability of an organic phase; and adding a potassium ionophore into the organic phase to form a selective ionic liquid colloid/water interface. In this way, the ionic liquid colloid/water interface with a high stability and a desirable selectivity is constructed. The interface is applied to the detection of K+ in a cerebral cortex, and is of great significance for studying a behavior of the K+ in vivo and a relationship of the K+ with diseases.

Abstract: A non-contact facial blood pressure measurement method based on 3D CNN is disclosed, which belongs to the technical field of computer vision. The method includes the following steps. S110: collecting an actual face video sample and training a blood pressure prediction model based on face images using 3D CNN neural network. S120: obtaining a face video in real time through a HD camera. S130: recognizing face key points in the face video obtained in S120 through dlib face recognition model, selecting a face region of interest, and extracting face images from the region. S140: performing a wavelet transform operation on the face images extracted in S130 to remove noise. S150: inputting seven consecutive frames of the face images into the 3D CNN blood pressure prediction model trained in S110 to obtain a blood pressure value of the measured person. The disclosure realizes non-contact facial blood pressure measurement.

Abstract: A method of stress compensation in a display device includes accumulating a brightness value of images displayed by the display device, encoding the accumulated brightness value to preserve a protected portion of the accumulated brightness value, compressing the accumulated brightness value, storing the compressed accumulated brightness value in a memory, retrieving and decompressing the compressed accumulated brightness value to generate a decompressed value, and decoding the decompressed value based on the protected portion to generate a decoded value corresponding to the accumulated brightness value, the decoded value having lower compression noise than the decompressed value.

Abstract: A method of stress compensation in a display device includes accumulating a brightness value of images displayed by the display device, encoding the accumulated brightness value to preserve a protected portion of the accumulated brightness value, compressing the accumulated brightness value, storing the compressed accumulated brightness value in a memory, retrieving and decompressing the compressed accumulated brightness value to generate a decompressed value, and decoding the decompressed value based on the protected portion to generate a decoded value corresponding to the accumulated brightness value, the decoded value having lower compression noise than the decompressed value.