Fine-tuning transformer models after unsupervised pre-training reaches a very high performance on many different natural language processing tasks. Unfortunately, transformers suffer from long inference times which greatly increases costs in production. One possible solution is to use knowledge distillation, solves this problem by transferring information large teacher smaller student models. K...

noise removal operation is commonly applied as pre-processing step before subsequent image processing tasks due to the occurrence of noise during acquisition or transmission process. a common problem in imaging systems by using cmos or ccd sensors is appearance of  the salt and pepper noise. this paper presents cellular automata (ca) framework for noise removal of distorted image by the salt an...

A composition procedure for linear and nondeleting extended top-down tree transducers is presented. It is demonstrated that the new procedure is more widely applicable than the existing methods. In general, the result of the composition is an extended top-down tree transducer that is no longer linear or nondeleting, but in a number of cases these properties can easily be recovered by a post-pro...

In this paper, a genetic algorithm is presented for an identical parallel-machine scheduling problem with family setup time that minimizes the total weighted flow time ( ). No set-up is necessary between jobs belonging to the same family. A set-up must be scheduled when switching from the processing of family i jobs to those of another family j, i  j, the duration of this set-up being the sequ...

In this paper, we will present a device specially designed on the basis of the theory of empty categories. This device cooperates with a bottom-up parser and is used as an elegant and efficient approachtotreatthetroublesome problems of the transformations of passivization,relativizatlon; toplcalization, ba-transformation and the use of zero pronouns in Chinese natural language. With the aid of ...

This paper addresses the issue of integrating top-down and bottom-up processing of information in modular hybrid systems. Backed by computational and neurological evidence, we show that there are important advantages to be gained from supporting bi-directional flow of information, especially in the context of intelligent information processing in real-world applications. We consider modular hyb...