Opis: This thesis explores the possibility of running neural networks on microcontrollers and how
to optimize their performance using instruction set extensions. Microcontrollers are seen as
too weak to run neural networks. We challenge this view and show that stripped-down neural
networks can run and be useful for some applications. We used an open-source microcontroller
called PULPino to run our neural network. The benefit of various instructions and optimizations
for minimizing energy consumption to run deep learning algorithms was evaluated. Hardware
loops, loop unrolling, and the dot-product unit were implemented and tested.
We developed an FPGA-based testing system to evaluate our hardware. We also developed a
deep learning library and a test neural network for our hardware. We wrote two versions of
the deep learning library. One version is the reference code, and the other is the optimized
code that uses the dot product unit. Using the testing system, we tested the performance of
the two versions. The synthesis was run to determine the power and energy consumption. We
also tried out various optimizations to see if the performance could be improved.
Using instruction set extensions and algorithmic optimizations we reduced the clock cycle count by 72% for the convolutional layers and by 78% for fully-connected layers. This reduced
power consumption by 73%. We compare our results with related research.Ključne besede: deep learning, embedded system, instruction set, RISC-VObjavljeno: 03.11.2020; Ogledov: 417; Prenosov: 83 Celotno besedilo (2,69 MB)
Opis: The article investigates attitudes towards English as a foreign language (EFL) by focusing on retrospective accounts of higher-education students' experience with learning English. The first part looks at individual factors affecting foreign language (FL) learning, such as attitudes towards FL learning and FL anxiety. The second part presents the results of a study conducted among students of the Faculty of Health Sciences at the University of Primorska. The main aim of the study was to identify students' attitudes towards their past EFL lessons and teachers and students' FL anxiety level. The results show that anxiety is a serious problem and that students exhibit alarmingly negative attitudes towards EFL lessons and teachers.Ključne besede: learning anxiety, foreign language anxiety, attitudes towards foreign language instruction, attitudes towards EFL teachers, English languageObjavljeno: 03.10.2017; Ogledov: 588; Prenosov: 83 Celotno besedilo (841,28 KB)Gradivo ima več datotek! Več...
Opis: In the last decade, educational researchers have been intensively searching for new, innovative teaching approaches. Information and Communication Technology (ICT) has a great didactic potential and project COLOS (Conceptual Learning of Science) encourages the use of ICT in the contemporary educational process. In this paper we present the conceptual learning of Physics. With experimental research we investigated the effectiveness of such learning in Slovenian secondary school. Two groups of third-year students who were enrolled in an introductory Physics course participated in the study. In the experimental group students were taught through the conceptual learning and in the control group a traditional expository instruction was used. We examined the knowledge of students after carrying out lessons specifically on the topic of Electricity. Five thinking processes were assessed - Knowledge (Recall), Analysis, Comparison, Inference and Evaluation. We found that the conceptual learning was more effective than the traditional instruction.Ključne besede: traditional instruction, conceptual learning, physics, ICT, simulations, physletsObjavljeno: 10.07.2015; Ogledov: 714; Prenosov: 275 Celotno besedilo (991,17 KB)Gradivo ima več datotek! Več...