Abstract: An information processing system includes: a face direction acquisition unit that obtains a face direction of a user; a gaze direction acquisition unit that obtains a gaze direction of the user; a determination unit that determines whether or not the user is a living body, on the basis of a difference between the face direction and the gaze direction, when an angle of the face direction is greater than or equal to a predetermined threshold; and an output unit that outputs a result of the determination. According to such an information processing system, it is possible to accurately determine whether or not the user is a living body.

Abstract: A computer-implemented method, a computer program product, and a computer processing system are provided for selecting from among multiple Graphics Processing Unit (GPU) execution modes for a Neural Network (NN) having a size greater than a threshold size. The multiple GPU execution modes include a normal memory mode, an Out-of-Core (OoC) execution mode, and a Unified Memory (UM) mode. The method includes starting an execution on the NN with the UM mode and measuring the memory usage for each of layers of the NN. The method further includes selecting an execution mode based on the memory usage of all of the layers.

Abstract: A method for converting a shape and a format of tensor data to meet a specific data format of a hardware accelerator is provided. The method receives input tensors L1 and L2, each being constants having a data format of < X x Y x Z >, and each further having an n-dimension input tensor shape as <Xn x Xn-1 x Xn-2 x ... x X1 >. The method stores input tensor shape. The method calculates an n-dimension modified shape of the input tensors by (a) setting a largest divisor of (Xn x Xn-1 x...x X1 ) ? L1 to S1, (b) setting a largest divisor of ((Xn x Xn-1 x...x X1 ) / S1) ? L2 to S2, (c) setting (((Xn x Xn-1 x... x X1 ) / (S1 x S2)) to S3, and (d) returning the n-dimension modified shape as < S3 x S2 x S1 >.

Abstract: A method is presented for compressing data of a Rectified Linear Unit (ReLU) function on a graphical processing unit (GPU) employed in a learning process of a deep neural network. The method includes converting an initial data structure including nonzero data and zero data into a compressed data structure including only the nonzero data of the initial data structure as compressed data by generating a nonzero data bitmap region, generating a nonzero data number table region by employing a parallel reduction algorithm, calculating a nonzero data array index per block region of all blocks from the nonzero data number table region by employing a parallel prefix sum scan algorithm, allocating a buffer for the compressed data; and copying the nonzero data from the initial data structure into a nonzero data array region in a compressed data format in parallel.

Abstract: Methods and systems for training a neural network include identifying units within a neural network, including a first unit for memory swapping and a second unit for re-computation to balance memory efficiency with computational efficiency. Each unit includes at least one layer of the neural network. Each unit has a first layer that is a checkpoint operation. During a feed-forward training stage, feature maps are stored in a first memory. The feature maps are output by the at least one layer of the first unit. The feature maps are swapped from the first memory to a second memory. During a backpropagation stage, the feature maps for the first unit are swapped from the second memory to the first memory. Feature maps for the second unit are re-computed.

Abstract: A method is presented for compressing data of a Rectified Linear Unit (ReLU) function on a graphical processing unit (GPU) employed in a learning process of a deep neural network. The method includes converting an initial data structure including nonzero data and zero data into a compressed data structure including only the nonzero data of the initial data structure as compressed data by generating a nonzero data bitmap region, generating a nonzero data number table region by employing a parallel reduction algorithm, calculating a nonzero data array index per block region of all blocks from the nonzero data number table region by employing a parallel prefix sum scan algorithm, allocating a buffer for the compressed data; and copying the nonzero data from the initial data structure into a nonzero data array region in a compressed data format in parallel.

Abstract: A computer-implemented method, computer program product, and computer processing system are provided for accelerating neural network data parallel training in multiple graphics processing units (GPUs) using at least one central processing unit (CPU). The method includes forming a set of chunks. Each of the chunks includes a respective group of neural network layers other than a last layer. The method further includes performing one or more chunk-wise synchronization operations during a backward phase of the neural network data parallel training, by each of the multiple GPUs and the at least one CPU.

Abstract: A computer-implemented method is provided for managing GPU memory consumption by computational graph rewriting. The method includes constructing, by a hardware processor, a categorized topological ordering of a computational graph. The categorized topological ordering includes multiple computational nodes arranged in multiple levels. The method further includes estimating, by the hardware processor, the GPU memory consumption responsive to a level including two or more computational nodes from among the multiple computational nodes. The method also includes rewriting, by the hardware processor, the computational graph by linearizing the two or more computational nodes in the level to avoid overlapping of the GPU memory consumption by the two or more computational nodes responsive to the GPU memory consumption exceeding a threshold. The memory additionally includes managing the GPU memory consumption in accordance with the rewritten computational graph.

Abstract: A method is provided for reducing consumption of a memory in a propagation process for a neural network (NN) having fixed structures for computation order and node data dependency. The memory includes memory segments for allocating to nodes. The method collects, in a NN training iteration, information for each node relating to an allocation, size, and lifetime thereof. The method chooses, responsive to the information, a first node having a maximum memory size relative to remaining nodes, and a second node non-overlapped with the first node lifetime. The method chooses another node non-overlapped with the first node lifetime, responsive to a sum of memory sizes of the second node and the other node not exceeding a first node memory size. The method reallocates a memory segment allocated to the first node to the second node and the other node to be reused by the second node and the other node.

Abstract: A computer-implemented method, a computer program product, and a computer processing system are provided for selecting from among multiple Graphics Processing Unit (GPU) execution modes for a Neural Network (NN) having a size greater than a threshold size. The multiple GPU execution modes include a normal memory mode, an Out-of-Core (OoC) execution mode, and a Unified Memory (UM) mode. The method includes starting an execution on the NN with the UM mode and measuring the memory usage for each of layers of the NN. The method further includes selecting an execution mode based on the memory usage of all of the layers.

Abstract: A method is provided for reducing consumption of a memory in a propagation process for a neural network (NN) having fixed structures for computation order and node data dependency. The memory includes memory segments for allocating to nodes. The method collects, in a NN training iteration, information for each node relating to an allocation, size, and lifetime thereof. The method chooses, responsive to the information, a first node having a maximum memory size relative to remaining nodes, and a second node non-overlapped with the first node lifetime. The method chooses another node non-overlapped with the first node lifetime, responsive to a sum of memory sizes of the second node and the other node not exceeding a first node memory size. The method reallocates a memory segment allocated to the first node to the second node and the other node to be reused by the second node and the other node.

Abstract: A computer-implemented method, computer program product, and computer processing system are provided for accelerating neural network data parallel training in multiple graphics processing units (GPUs) using at least one central processing unit (CPU). The method includes forming a set of chunks. Each of the chunks includes a respective group of neural network layers other than a last layer. The method further includes performing one or more chunk-wise synchronization operations during a backward phase of the neural network data parallel training, by each of the multiple GPUs and the at least one CPU.

Abstract: An extraction method for extracting a sub query to be converted to a program for processing stream data continuously inputted to a database, from a query including instructions, as sub queries, to be issued to a database management system. The extraction method includes receiving input of the query and a lower limit value of efficiency as processing time per unit memory increase amount. A calculating operation calculates a one memory increase amount and the efficiency using the memory increase amount and the processing time to be reduced. The method selects a sub query whose calculated efficiency is equal to or higher than the lower limit value and extracts the selected sub query as a conversion object on condition that the integrated memory increase amount is equal to or smaller than the maximum memory increase amount.

Abstract: An extraction device for extracting a sub query to be converted to a program for processing stream data continuously inputted to a database, from a query including instructions, as sub queries, to be issued to a database management system. The extraction device includes: an input unit; an operation unit for calculating the memory increase amount in a case of processing the stream data and the processing time to be reduced for each sub query, and calculating the efficiency by using them; and an extraction unit for selecting at least one sub query whose efficiency is equal to or higher than the lower limit value, integrating the memory increase amount calculated for the selected sub query, and on condition that the integrated memory increase amount is equal to or smaller than the maximum memory increase amount, extracting the selected sub query as a conversion object.

Abstract: An extraction method for extracting a sub query to be converted to a program for processing stream data continuously inputted to a database, from a query including instructions, as sub queries, to be issued to a database management system. The extraction method includes receiving input of the query and a lower limit value of efficiency as processing time per unit memory increase amount. A calculating operation calculates a one memory increase amount and the efficiency using the memory increase amount and the processing time to be reduced. The method selects a sub query whose calculated efficiency is equal to or higher than the lower limit value and extracts the selected sub query as a conversion object on condition that the integrated memory increase amount is equal to or smaller than the maximum memory increase amount.

Abstract: An extraction device for extracting a sub query to be converted to a program for processing stream data continuously inputted to a database, from a query including instructions, as sub queries, to be issued to a database management system. The extraction device includes: an input unit; an operation unit for calculating the memory increase amount in a case of processing the stream data and the processing time to be reduced for each sub query, and calculating the efficiency by using them; and an extraction unit for selecting at least one sub query whose efficiency is equal to or higher than the lower limit value, integrating the memory increase amount calculated for the selected sub query, and on condition that the integrated memory increase amount is equal to or smaller than the maximum memory increase amount, extracting the selected sub query as a conversion object.