Abstract: Systems and methods for multi-modality anomaly detection using artificial intelligence models such as fused models. Metric data and log data obtained from a monitored entity can be encoded into metric representations and log representations by utilizing transformer encoders of a cross-joint variational autoencoder (CJVAE). The metric representations and the log representations can be fused into a joint context representation by utilizing a fusion transformer encoder of the CJVAE. The joint context representation can be decoded by utilizing transformer decoders of the CJVAE to reconstruct the metric representations and the log representations. An anomaly for the monitored entity can be detected by aggregating detection results from the CJVAE based on the metric representations and the log representations, a metric-specific detection result from a metric detector, and a log-specific detection result from a log detector to resolve determined issues of the monitored entity caused by the anomaly.

Abstract: Caching write input/output (I/O) operations in a replica-based storage system. A write I/O operation is received from a consumer, and a corresponding replica list is identified. A first replica set is selected from the replica list for caching the I/O operation, and a first log corresponding to the I/O operation is added to a primary ring buffer of the first replica set. When the first log cannot be replicated to a secondary ring buffer of the first replica set, a second replica set is selected from the replica list for caching the I/O operation. A second log corresponding to the I/O operation is added to a primary ring buffer of the second replica set. Once the second log has been replicated to a secondary ring buffer of the second replica set, the I/O operation is acknowledged to the consumer, and the second log is de-staged to a backing store.

Abstract: Methods and systems include generating general annotations for input time series data based on annotations from one or more source domains. Domain-specific annotations are generated for the input time series based on annotations from a target domain and based on the general annotations. An action is performed responsive to the domain-specific annotations and the general annotations.

Abstract: A system comprises a first computer system and a second computer system. The first computer system includes a processor and a computer-readable medium storing instructions executable to determine that a primary host in a replica set is unavailable, the replica set comprising a primary ring buffer and one or more secondary ring buffers stored across a plurality of hosts. The first computer system is further configured to choose a secondary host as a de-stage primary for the replica set and to communicate an election to the secondary host. The second computer system includes a processor and a computer-readable medium storing instructions executable to receive the election as the de-stage primary for the replica set, identify a ring buffer stored in persistent memory comprising logs replicated from the primary ring buffer, and de-stage the logs from the ring buffer to a backing store.

Abstract: Methods and systems for transitioning write input/output (I/O) requests from a write-caching mode to a pass-through mode. The method includes determining when a condition has been met for transitioning the write I/O requests and de-staging one or more logs from a write cache to a backing store, where each log corresponds to a pending write I/O operation by a consumer. The method further includes determining when no log remains in the write cache for de-staging to the backing store and initiating the pass-through mode by routing new write I/O requests directly to the backing store instead of the write cache.

Abstract: Requests for a write storage operation are stored in a ring buffer. The next available request is retrieved from the ring buffer. When the next available request comprises a write operation that overlaps in a logical address range with a previous request stored in the ring buffer, the next available request is suspended. The next available request and the previous request are merged to generate a single request that includes logical address ranges of the next available request and the previous request.

Abstract: Requests for a write storage operation are stored in a ring buffer. The write storage operations are executed using polling threads and cache de-stage threads. Dispatchers and worker threads are created for executing the polling threads and cache de-stage threads. Queue pairs for each pair of dispatchers and worker threads are generated. The queue pairs comprise a submission queue and a completion queue. The next available request is retrieved from the ring buffer. A scoring algorithm is used to load balance the queue pairs associated with the worker threads, the scoring algorithm operable to determine a score based a current depth of the submission queue and completion queue.

Abstract: Requests for a write storage operation are stored in a ring buffer. The next available request is retrieved from the ring buffer. When the next available request comprises a write operation that overlaps in a logical address range with a previous request stored in the ring buffer, the next available request is suspended. The next available request and the previous request are merged to generate a single request that includes logical address ranges of the next available request and the previous request.

Abstract: Systems and methods for domain adaptation of artificial intelligence (AI) models based on multi-source time-series data. Meta-data information from tuples of time-series data and corresponding labels for the time-series data can be learned based on a fidelity loss with a prompt-based deep learning model (POND) using determined soft prompts. Mutual information from the meta-data information can be minimized by minimizing a discrimination loss of domain-specific information from the meta-data information. A common prompt can be learned with a learning objective that combines a training loss, the fidelity loss and the discrimination loss. AI models can be adapted to perform downstream tasks for different domains by utilizing the common prompt for the AI models.

Abstract: A computer system and method are disclosed for replicating logs in a distributed environment. The method includes identifying a write input/output (I/O) operation and identifying a log to be replicated based on the write I/O operation. The log is then persisted to local non-volatile memory in the computer system. Subsequently, the log is replicated to multiple remote hosts, where each remote host of the plurality of remote hosts stores the log in its corresponding local non-volatile memory without de-staging the log to a backing store. The write I/O operation is committed once the log is replicated to at least a subset of the remote hosts that forms a quorum. Finally, the log is de-staged to the backing store after the write I/O operation is successfully committed.

Abstract: Methods and systems for transitioning write input/output (I/O) requests from a write-caching mode to a pass-through mode. The method includes determining when a condition has been met for transitioning the write I/O requests and de-staging one or more logs from a write cache to a backing store, where each log corresponds to a pending write I/O operation by a consumer. The method further includes determining when no log remains in the write cache for de-staging to the backing store and initiating the pass-through mode by routing new write I/O requests directly to the backing store instead of the write cache.

Abstract: Caching write input/output (I/O) operations in a replica-based storage system. A write I/O operation is received from a consumer, and a corresponding replica list is identified. A first replica set is selected from the replica list for caching the I/O operation, and a first log corresponding to the I/O operation is added to a primary ring buffer of the first replica set. When the first log cannot be replicated to a secondary ring buffer of the first replica set, a second replica set is selected from the replica list for caching the I/O operation. A second log corresponding to the I/O operation is added to a primary ring buffer of the second replica set. Once the second log has been replicated to a secondary ring buffer of the second replica set, the I/O operation is acknowledged to the consumer, and the second log is de-staged to a backing store.

Abstract: Methods and systems include adapting an initial prompt to a target domain corresponding to an input time series to generate an adapted prompt. The adapted prompt and the input time series are combined. The input time series is processed with the adapted prompt using a modular transformer encoder that has a plurality of sub-encoders, with a policy network selecting a subset of the plurality of encoders that are applied to the input time series and the adapted prompt.

Abstract: Systems and methods are provided for adapting a model trained from multiple source time-series domains to a target time-series domain, including integrating input data from source time-series domains to pretrain a model with a set of domain-invariant representations, fine-tuning the model by learning prompts specific to each source time-series domain using data from the source time-series domains, and applying instance normalization and segmenting the time-series data into subseries-level normalized patches for the target time-series domain. The normalized patches are fed into a transformer encoder to generate high-dimensional representations of the normalized patches, and a limited number of samples from the target time-series domain are utilized to learn the prompt specific to the target domain.

Abstract: A method for generating mark information in a virtual environment includes displaying a target-perspective picture showing the virtual environment, a sight bead pointing to an aiming position, and receiving a mark operation triggered on a mark control interface. The method further includes, in response to receiving the mark operation, generating, according to the received mark operation, mark information on a virtual object to which the aiming position points. The virtual object on which the mark information is generated is a movable object, a virtual prop, a virtual building, or a virtual landscape object in the virtual environment. The generated mark information is visible at a position of the virtual object in the virtual environment to one or more teammates of the virtual character.

Abstract: A tunnel dynamic blasting device based on geological body intelligent perception, a system and a method are provided. A blasting design instrument body of the tunnel dynamic blasting device is disposed with a 3D laser scanning device, a digital camera device, a borehole locating device, an electronic screen, a built-in computer and a wireless communication device. The 3D laser scanning device, the digital camera device, the borehole locating device, the electronic screen and the wireless communication device are connected to the built-in computer. A housing of the blasting design instrument body is disposed with a battery compartment having a lithium battery secured therein. A range finder is secured above the housing. Problems of non-systematicness and lag of tunnel blasting design and coarseness of borehole locating nowadays are solved, thus the tunnel excavation efficiency and borehole arrangement accuracy in the tunnel cross-section are improved, thereby improving the tunnel blasting quality.

Abstract: A computer-implemented method is provided for cross-lingual transfer. The method includes randomly masking a source corpus and a target corpus to obtain a masked source corpus and a masked target corpus. The method further includes tokenizing, by pretrained Natural Language Processing (NLP) models, the masked source corpus and the masked target corpus to obtain source tokens and target tokens. The method also includes transforming the source tokens and the target tokens into a source dependency parsing tree and a target dependency parsing tree. The method additionally includes inputting the source dependency parsing tree and the target dependency parsing tree into a graph encoder pretrained on a translation language modeling task to extract common language information for transfer. The method further includes fine-tuning the graph encoder and a down-stream network for a specific NLP down-stream task.

Abstract: In a method for displaying a search image, a virtual scene from a perspective of a virtual object that is in flight and a user interface (UI) are displayed. The search image is displayed with the UI in response to a continuous touch operation. The search image is from a search perspective of the virtual object and corresponds to a local region of the virtual scene. The display of the search image is updated in response to an adjustment of the search perspective according to the continuous touch operation.

Abstract: A method for generating mark information in a virtual environment includes displaying a target-perspective picture showing the virtual environment, a sight bead pointing to an aiming position, and receiving a mark operation triggered on a mark control interface. The method further includes, in response to receiving the mark operation, generating, according to the received mark operation, mark information on a virtual object to which the aiming position points. The virtual object on which the mark information is generated is a movable object, a virtual prop, a virtual building, or a virtual landscape object in the virtual environment. The generated mark information is visible at a position of the virtual object in the virtual environment to one or more teammates of the virtual character.

Abstract: A method for generating mark information in a virtual environment includes displaying a target-perspective picture showing the virtual environment including a virtual character from a perspective, a sight bead and a mark control interface being displayed in the target-perspective picture. The method also includes receiving a first-perspective switching operation, and changing an observation direction of the target-perspective picture and an aiming position of the sight bead according to the first-perspective switching operation. The method further includes receiving a mark operation triggered on the mark control interface, and generating, according to the mark operation, mark information on a virtual object to which the aiming position points.