- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[IEEE TENCON 2019 - 2019 IEEE Region 10 Conference (TENCON) - Kochi, India (2019.10.17-2019.10.20)] TENCON 2019 - 2019 IEEE Region 10 Conference (TENCON) - Speech Enabled Visual Question Answering using LSTM and CNN with Real Time Image Capturing for assisting the Visually Impaired
摘要： The proposed work benefits visually impaired individuals in identifying objects and visualizing scenarios around them independent of any external support. In such a situation, the surrounding and ask an open-ended question, classification question, counting question or yes/no question to the application by speech input. The proposed application uses Visual Question Answering (VQA) to integrate image processing and natural language processing which is also capable of speech to text translation and vice versa that helps to identify, recognize and thus obtain details of any particular image. The work uses a classical CNN-LSTM model where image features and language features are computed separately and combined at a later stage using image features and word embedding obtained from the question and runs a multilayer perceptron on the combined features to obtain the results. The model achieves an accuracy of 57 per cent. The model can also be utilized to develop cognitive interpretation better in kids. As the application is speech enabled it is best suited for the visually impaired with an easy to use GUI.
关键词： VGG16,Visually Impaired,Keras Neural Network Library,ImageNet,gTTS,Feature extraction,Image Recognition,VQA,Word2Vec,Speech Recognition,Glove vector,CNN,Multi Layer Perceptron,LSTM
更新于2025-09-16 10:30:52
-
Sensor Fusion for Distance Estimation Under Disturbance with Reflective Optical Sensors using Multi Layer Perceptron (MLP)
摘要： There are many methods to perform distance measurement, among them the reflexive optical sensors, which are low cost but present some issues such as nonlinear response, limited operating ranges and the measurement shows sensitivity to infrared or visible radiation. For this reason, this work presents a sensory fusion model combining three reflective optical distance sensors of different ranges, a color sensor (VIS), an ultraviolet radiation sensor (UV) and an Near infrared sensor (NIR) to estimate the distance using a Multi Layer Perceptron (MLP). The purpose of combining different distance sensors is to have a higher overall range and achieve redundancy in some regions, and the objective of the UV-VIS-NIR sensors is to compensate for the radiation at which the distance sensors are exposed to adjust the measurement. With the attained information, the influence of each type of radiation on the distance measurement was evaluated. It is important to estimate the distance in these ranges because in robotics and automation industries, different associated applications are handled. The MLP was trained switching its architecture between four and sixteen neurons per layer, and between three and five hidden layers. Finally the training and selection of several MLP architectures for sensory fusion with an error lower than 1% was presented.
关键词： Sensor fusion,Distance estimation,Reflective optical sensor,Infrarred sensor,Multi Layer Perceptron (MLP)
更新于2025-09-16 10:30:52
-
[IEEE 2019 IEEE/ACM Workshop on Photonics-Optics Technology Oriented Networking, Information and Computing Systems (PHOTONICS) - Denver, CO, USA (2019.11.18-2019.11.18)] 2019 IEEE/ACM Workshop on Photonics-Optics Technology Oriented Networking, Information and Computing Systems (PHOTONICS) - An Optical Neural Network Architecture based on Highly Parallelized WDM-Multiplier-Accumulator
摘要： Future applications such as anomaly detection in a network and autonomous driving require extremely low, sub-microsecond latency processing in pattern classification. Towards the realization of such an ultra-fast inference processing, this paper proposes an optical neural network architecture which can classify anomaly patterns at sub-nanosecond latency. The architecture fully exploits optical parallelism of lights using wavelength division multiplexing (WDM) in vector-matrix multiplication. It also exploits a linear optics with passive nanophotonic devices such as microring resonators, optical combiners, and passive couplers, which make it possible to construct low power and ultra-low latency optical neural networks. Optoelectronic circuit simulation using optical circuit implementation of multi-layer perceptron (MLP) demonstrates sub-nanosecond processing of optical neural network.
关键词： optical neural network,wavelength division multiplexing,multi-layer perceptron
更新于2025-09-16 10:30:52
-
Predicting the water production of a solar seawater greenhouse desalination unit using multi-layer perceptron model
摘要： A solar seawater greenhouse is a type of desalination plant that uses solar energy and seawater to humidify the interior of the greenhouse and produce fresh water from the humid air (humidification-dehumidification process). The produced water is used both for irrigating agricultural crops and for drinking. Many parameters affect the performance of seawater greenhouses. The present study employed an artificial neural network to examine the effective parameters of the greenhouse on the fresh water production such as width, length, the height of the front evaporator, and roof transparency. A suitable structure was obtained for the multi-layer perceptron (MLP) method and the mathematical statistics % AARE, RMSE, and R2, were used to evaluate network performance. The method showed good agreement with the experimental data. Using the optimized created network, the effect of each parameter on the produced fresh water was assessed. Finally, a 125 m wide and 200 m long of greenhouse with a 4 m height of front evaporator and roof transparency of 0.6 that produced 161.6 m3/day of fresh water was introduced as the optimal seawater greenhouse.
关键词： Solar seawater greenhouse,Desalination,Humidification-dehumidification process,Multi-layer perceptron
更新于2025-09-09 09:28:46