Abstract: The invention discloses a Schleiferilactobacillus harbinensis (former name Lactobacillus harbinensis) M1 and use thereof. The preservation number of Schleiferilactobacillus harbinensis M1 is GDMCC No. 60305. The strain is sensitive to antibiotics such as ampicillin, tetracycline and chloramphenicol, and its safety meets the requirements of EFSA. Its supernatant has a significant inhibitory effect on Listeria monocytogenes, Streptococcus hemolyticus and Staphylococcus aureus. The present invention also protects the use of Schleiferilactobacillus harbinensis M1 in fermented soymilk; using the bacteria to ferment soymilk alone has the advantages of strong acid production ability, high amount of viable bacteria, few or even completely no beany flavor substances, and significant higher content of characteristic aroma components of fermented milk, which greatly improves the sensory flavor of fermented soymilk.

Abstract: The present disclosure is directed to a microbiostatic cleansing composition for use on skin and hard surfaces that reduces microbe resistance. The cleansing composition may be, for example, bacteriostatic or virostatic and includes a release agent that entangles the microbes without killing them allowing them to be removed from skin and hard surface in an inactive state.

Abstract: This application provides a method for constructing a network structure optimizer performed by an electronic device.

Abstract: In some embodiments, a method trains a first prediction network to predict similarity between images in videos. The training uses boundaries detected in the videos to train the prediction network to predict images in a same scene to have similar feature descriptors. The first prediction network generates feature descriptors that describe library images from videos in a video library offered to users of a video delivery service. A search image is received and the prediction network predicts one or more library images for one or more videos that are predicted to be similar to the received image. The one or more library images for the one or more videos are provided as a search result.

Abstract: A method receives ratings for videos from a first user that is using a video delivery service. A first model includes connection networks where each connection network corresponds to a rating. The method inputs each rating into a connection network in an order. Also, parameters for the ratings and ratings other than the rating received from the first user are modeled in a respective connection network. Values for the set of parameters are trained such that the plurality of connection networks predict conditional probabilities that the first user would provide the rating corresponding to the each connection network in the order. The conditional probabilities are based on the first user providing ratings that are previously located in the order. The parameters are then used to generate a list of videos to recommend to the first user using the first model.

Abstract: In some embodiments, a method trains a first prediction network to predict similarity between images in videos. The training uses boundaries detected in the videos to train the prediction network to predict images in a same scene to have similar feature descriptors. The first prediction network generates feature descriptors that describe library images from videos in a video library offered to users of a video delivery service. A search image is received and the prediction network predicts one or more library images for one or more videos that are predicted to be similar to the received image. The one or more library images for the one or more videos are provided as a search result.

Abstract: A method receives ratings for videos from a first user that is using a video delivery service. A first model includes connection networks where each connection network corresponds to a rating. The method inputs each rating into a connection network in an order. Also, parameters for the ratings and ratings other than the rating received from the first user are modeled in a respective connection network. Values for the set of parameters are trained such that the plurality of connection networks predict conditional probabilities that the first user would provide the rating corresponding to the each connection network in the order. The conditional probabilities are based on the first user providing ratings that are previously located in the order. The parameters are then used to generate a list of videos to recommend to the first user using the first model.

Abstract: Provided herein is an eyeglasses frame, wherein one optical frame which can fix a lens and has two fixed segments is provided, one connecting segment of an eyeglasses leg is provided on the optical frame adjacent to each fixed segment, respectively, one end of the eyeglasses leg extends from a frame side to another end of the eyeglasses legs far away from the frame, the eyeglasses leg is fixed on the connecting segment, can be inverted, extends by being basically parallel to the optical frame at a first end and extends backwards from the optical frame at a second frame. Also provided herein is and optical frame used in the eyeglasses frame and a mounting method for the eyeglasses frame and eyeglasses.