Abstract: An embossment (100) on a sheet of film and a glove produced from the embossed sheet of film is disclosed. The embossment (100) comprises a base surface (103); a raised surface (102) opposite to the base surface (103); and at least one recessed area (104) formed therein and extending from the base surface (103) and the raised surface (102) to the base surface (103), wherein depth (Dr) of the recessed area (104) is between 50% to 99% relative to height (He) of the embossment (100) between the base surface (103) and the raised surface (102).

Abstract: A method of treating a subject suffering from cancer comprising a step of administering an effective amount of a group of double-stranded RNA molecules to the subject, wherein the RNA molecule is isolated or derived from a bacteria of the genus Escherichia. A method of inhibiting growth or proliferation of cancer cells comprising a step of contacting said cells with said RNA molecule; and a pharmaceutical composition for treating cancer comprising said RNA molecule and a pharmaceutically tolerable excipient. Also a double-stranded RNA molecule and a recombinant vector comprising the double-stranded RNA molecule.

Abstract: The present disclosure proposes a fixed-point training method and apparatus based on static fixed-point conversion scheme. More specifically, the present disclosure proposes a fixed-point training method for LSTM neural network. According to this method, during the fine-tuning process of the neural network, it uses fixed-point numbers to conduct forward calculation. Accordingly, within several training cycles, the network accuracy may returned to the desired accuracy level under floating point calculation.

Abstract: A method of treating a subject suffering from cancer includes administering an effective amount of a RNA molecule to the subject, wherein the RNA molecule is isolated or derived from a plant of the genus Taxus. A method of inhibiting growth or proliferation of cancer cells includes contacting cancer cells with the RNA molecule; and a pharmaceutical composition for treating cancer includes the RNA molecule and a pharmaceutically tolerable excipient. Also a double-stranded RNA molecule and a recombinant vector include the double-stranded RNA molecule.

Abstract: A method of treating a subject suffering from cancer comprising a step of administering an effective amount of a group of double-stranded RNA molecules to the subject, wherein the RNA molecule is isolated or derived from a bacteria of the genus Escherichia. A method of inhibiting growth or proliferation of cancer cells comprising a step of contacting said cells with said RNA molecule; and a pharmaceutical composition for treating cancer comprising said RNA molecule and a pharmaceutically tolerable excipient. Also a double-stranded RNA molecule and a recombinant vector comprising the double-stranded RNA molecule.

Abstract: Application of transfer RNA Molecules and their derived fragments for prevention or treatment of heart disease. The present invention provides a method of preventing or treating a subject suffering from heart diseases comprising administration of transfer RNA molecules and fragments derived from transfer RNA molecules or its functional variants or homologous to the subject, wherein the RNA molecules isolated from or derived from a plant of the genus Panax. The present invention also provides a pharmaceutical composition for the prevention or treatment of heart diseases comprising said effective amount of RNA molecule and a pharmaceutically tolerable vector, virus or excipient. The present invention provides a method for the prevention or treatment of a subject suffering from a heart disease. It is found that transfer RNA molecules from ginseng are particularly effective in the treatment of heart diseases, and also have a restorative effect on the myocardial cytoskeleton after ischemia-reperfusion injury.

Abstract: The present disclosure proposes a fixed-point training method and apparatus based on dynamic fixed-point conversion scheme. More specifically, the present disclosure proposes a fixed-point training method for LSTM neural network. According to this method, during the fine-tuning process of the neural network, it uses fixed-point numbers to conduct forward calculation. Accordingly, within several training cycles, the network accuracy may returned to the desired accuracy level under floating point calculation.

Abstract: A method of treating a subject suffering from cancer includes administering an effective amount of a RNA molecule to the subject, wherein the RNA molecule is isolated or derived from a plant of the genus Taxus. A method of inhibiting growth or proliferation of cancer cells includes contacting cancer cells with the RNA molecule; and a pharmaceutical composition for treating cancer includes the RNA molecule and a pharmaceutically tolerable excipient. Also a double-stranded RNA molecule and a recombinant vector include the double-stranded RNA molecule.

Abstract: The present disclosure proposes an improved compression method for neural networks (e.g. LSTM), which may effectively shorten the training period of a neural network by combining pruning operation into the training process, so as to reduce the number of iteration in the training process.

Abstract: The present disclosure proposes a fixed-point training method and apparatus based on dynamic fixed-point conversion scheme. More specifically, the present disclosure proposes a fixed-point training method for LSTM neural network. According to this method, during the fine-tuning process of the neural network, it uses fixed-point numbers to conduct forward calculation. Accordingly, within several training cycles, the network accuracy may returned to the desired accuracy level under floating point calculation.

Abstract: The present disclosure proposes a fixed-point training method and apparatus based on static fixed-point conversion scheme. More specifically, the present disclosure proposes a fixed-point training method for LSTM neural network. According to this method, during the fine-tuning process of the neural network, it uses fixed-point numbers to conduct forward calculation. Accordingly, within several training cycles, the network accuracy may returned to the desired accuracy level under floating point calculation.