CNN has shown excellent performance on object recognition based on huge amount of real images. For training with synthetic data rendered from 3D models alone to reduce the workload of collecting real images, we propos...
详细信息
ISBN:
(纸本)9781467399623
CNN has shown excellent performance on object recognition based on huge amount of real images. For training with synthetic data rendered from 3D models alone to reduce the workload of collecting real images, we propose a concatenated self-restraint learning structure lead by a triplet and softmax jointed loss function for object recognition. Locally connected auto encoder trained from rendered images with and without background used for object reconstruction against environment variables produces an additional channel automatically concatenated to RGB channels as input of classification network. This structure makes it possible training a softmax classifier directly from CNN based on synthetic data with our rendering strategy. Our structure halves the gap between training based on real photos and 3D model in both PASCAL and ImageNet database compared to GoogleNet.
This work analyzes centered Restricted Boltzmann Machines (RBMs) and centered Deep Boltzmann Machines (DBMs), where centering is done by subtracting offset values from visible and hidden variables. We show analyticall...
详细信息
This work analyzes centered Restricted Boltzmann Machines (RBMs) and centered Deep Boltzmann Machines (DBMs), where centering is done by subtracting offset values from visible and hidden variables. We show analytically that (i) centered and normal Boltzmann Machines (BMs) and thus RBMs and DBMs are different parameterizations of the same model class, such that any normal BM/RBM/DBM can be transformed to an equivalent centered BM/RBM/DBM and vice versa, and that this equivalence generalizes to artificial neural networks in general, (ii) the expected performance of centered binary BMs/RBMs/DBMs is invariant under simultaneous flip of data and offsets, for any off-set value in the range of zero to one, (iii) centering can be reformulated as a different update rule for normal BMs/RBMs/DBMs, and (iv) using the enhanced gradient is equivalent to setting the offset values to the average over model and data mean. Furthermore, we present numerical simulations suggesting that (i) optimal generative performance is achieved by subtracting mean values from visible as well as hidden variables, (ii) centered binary RBMs/DBMs reach significantly higher log-likelihood values than normal binary RBMs/DBMs, (iii) centering variants whose offsets depend on the model mean, like the enhanced gradient, suffer from severe divergence problems, (iv) learning is stabilized if an exponentially moving average over the batch means is used for the offset values instead of the current batch mean, which also prevents the enhanced gradient from severe divergence, (v) on a similar level of log-likelihood values centered binary RBMs/DBMs have smaller weights and bigger bias parameters than normal binary RBMs/DBMs, (vi) centering leads to an update direction that is closer to the natural gradient, which is extremely efficient for training as we show for small binary RBMs, (vii) centering eliminates the need for greedy layer-wise pre-training of DBMs, which often even deteriorates the results independent
This paper proposes a sensor fault detection system for a two-level DVR, controlled by a repetitive controller. The system compensates key voltage-quality disturbances namely;voltage sags, harmonic voltages, voltage i...
详细信息
ISBN:
(纸本)9781467361538
This paper proposes a sensor fault detection system for a two-level DVR, controlled by a repetitive controller. The system compensates key voltage-quality disturbances namely;voltage sags, harmonic voltages, voltage imbalances, and control current during downstream fault, additionally detect any fault in sensor measurements as well. All the control actions of controller depend on the availability and quality of sensor measurement. However, measurements are inevitably subjected to faults caused by sensor failure, broken or bad connections, bad communication, or malfunction of some hardware or software. Therefore an auto-associative neural network based system is used here to detect any fault in sensor measurement. MATLAB/SIMULINK is used to carry out all modeling aspects of test system.
This paper proposes a sensor fault detection system for a two-level DVR, controlled by a repetitive controller. The system compensates key voltage-quality disturbances namely;voltage sags, harmonic voltages, voltage i...
详细信息
ISBN:
(纸本)9781467361521
This paper proposes a sensor fault detection system for a two-level DVR, controlled by a repetitive controller. The system compensates key voltage-quality disturbances namely;voltage sags, harmonic voltages, voltage imbalances, and control current during downstream fault, additionally detect any fault in sensor measurements as well. All the control actions of controller depend on the availability and quality of sensor measurement. However, measurements are inevitably subjected to faults caused by sensor failure, broken or bad connections, bad communication, or malfunction of some hardware or software. Therefore an auto-associative neural network based system is used here to detect any fault in sensor measurement. MATLAB/SIMULINK is used to carry out all modeling aspects of test system.
Question generation is an important task in natural language processing that involves generating questions from a given text. This paper proposes a novel approach for dynamic question generation using a context-aware ...
详细信息
Question generation is an important task in natural language processing that involves generating questions from a given text. This paper proposes a novel approach for dynamic question generation using a context-aware auto-encoded graph neural model. Our approach involves constructing a graph representation of the input text, where each node in the graph corresponds to a word or phrase in the text, and the edges represent the relationships between them. We then use an auto-encoder model to learn a compressed representation of the graph that captures the most important information in the input text. Finally, we use the compressed graph representation to generate questions by dynamically selecting nodes and edges based on their relevance to the context of the input text. We evaluate our approach on four benchmark datasets (SQuAD, Natural Questions, TriviaQA, and QuAC) and demonstrate that it outperforms existing state-of-the-art methods for dynamic question generation. In the experimentation, to evaluate the result four performance metrics are used i.e. BLEU, ROUGE, F1-Score, and Accuracy. The result of the proposed approach yields an accuracy of 92% on the SQuAD dataset, 89% with QuAC, and 84% with TriviaQA. while on the natural questions dataset, the model gives 79% accuracy. Our results suggest that the use of graph neural networks and auto-encoder models can significantly improve the accuracy and effectiveness of question generation in NLP. Further research in this area can lead to even more sophisticated models that can generate questions that are even more contextually relevant and natural-sounding.
暂无评论