TL;DR: Fastai as mentioned in this paper is a deep learning library which provides practitioners with high-level components that can quickly and easily provide state-of-the-art results in standard deep learning domains, and provides researchers with low-level component that can be mixed and matched to build new approaches.
Abstract: fastai is a deep learning library which provides practitioners with high-level components that can quickly and easily provide state-of-the-art results in standard deep learning domains, and provides researchers with low-level components that can be mixed and matched to build new approaches. It aims to do both things without substantial compromises in ease of use, flexibility, or performance. This is possible thanks to a carefully layered architecture, which expresses common underlying patterns of many deep learning and data processing techniques in terms of decoupled abstractions. These abstractions can be expressed concisely and clearly by leveraging the dynamism of the underlying Python language and the flexibility of the PyTorch library. fastai includes: a new type dispatch system for Python along with a semantic type hierarchy for tensors; a GPU-optimized computer vision library which can be extended in pure Python; an optimizer which refactors out the common functionality of modern optimizers into two basic pieces, allowing optimization algorithms to be implemented in 4-5 lines of code; a novel 2-way callback system that can access any part of the data, model, or optimizer and change it at any point during training; a new data block API; and much more. We have used this library to successfully create a complete deep learning course, which we were able to write more quickly than using previous approaches, and the code was more clear. The library is already in wide use in research, industry, and teaching. NB: This paper covers fastai v2, which is currently in pre-release at this http URL
TL;DR: In this paper, a callback system is created utilizing a hybrid telecommunication system including a switched communication network and a packet transmission network, where a call parameter database is stored in a memory.
Abstract: A callback system is created utilizing a hybrid telecommunication system including a switched communication network and a packet transmission network. A call parameter database is stored in a memory. A call is received on the system. The call parameter database is accessed to determine at least one call parameter. The call is routed over the switched communication network and the packet transmission network based on the at least one call parameter. A plurality of service engines is provided, each configured to execute desired service logic utilizing expert system.
TL;DR: In this article, the Louth Automation video disk communications protocol supports continuous play over multiple clips for extended periods of time, allows a playlist to be edited dynamically after being given to the video server and during playback of clips in the play-list, allows some notion of "current time" to be used during the streaming of continuous media data, and supports features of the "Louth automation" video disk communication protocol.
Abstract: A protocol and interface provides continuous play over multiple clips for extended periods of time, allows a play-list to be edited dynamically after being given to the video server and during playback of clips in the play-list, allows some notion of "current time" to be used during the streaming of continuous media data, and supports features of the "Louth Automation" video disk communications protocol. Preferably, the client application first creates a session with a play-list containing a fixed number of entries; the number should be as small as possible consistent with the client's requirements. The client edits this play-list by appending the first few clips and then starts the session playing. Each time transmission of video data of a clip is completed, the clip is removed from the head of the play-list, all other clips are moved down, and a callback is issued to the client with the current, updated, play-list. A callback is also issued with the updated play-list to acknowledge each edit command. Preferably, there is a limit as to how close to air-time a clip normally may be deleted or new material inserted, in order to ensure continuity of transmission of the video stream of each clip. To allow live break-ins or other "emergency" operations, however, the session may be paused and later resumed and subsequent clips may be "trimmed" to reduce their play times to recover the time lost to the break-in.
TL;DR: In this paper, a call queueing system for a call center establishes virtual calls for callers, and allows the callers to disconnect and await a later callback when the virtual call is routed to an agent.
Abstract: A call queueing system for a call center establishes virtual calls for callers, and allows the callers to disconnect and await a later callback when the virtual call is routed to an agent. Callers are connected to an IVR and informed that they may disconnect and await the callback. An IVR may also be used when the callback is made, informing the caller of the nature of the new call, and that they are now being connected to an agent. In some instances virtual calls may be sequenced in the same queue with real calls. The virtual call system may be initiated in a call center by exceeding load threshold, at particular time frames, and the like. A key is associated with each virtual call in a manner that the origination number of the call for whom the virtual call is entered may be retrieved when the call is routed to an agent. The system is adapted to work with all intelligent routing systems.
TL;DR: In this paper, a system and method of answering incoming calls are disclosed in which, if the resource (18)(19) desired by the caller is not then available, a robot controller (14) will inform the caller when a call back to the caller can be made.
Abstract: A system and method of answering incoming calls are disclosed in which, if the resource (18)(19) desired by the caller is not then available, a robot controller (14) will inform the caller when a call back to the caller can be made. The callback time can be suggested by the robot controller (14) or the caller can request a specific callback time. The robot controller (14) ascertains and verifies the caller's call back identity and stores the number of the identity in a callback queue (11) along with such other pertinent information as the callback time, the caller's name and the required resource. At the appropriate time, the robot controller (14) removes the information from the callback queue (11), places the call and connects the desired resource (18)(19).