Multi-attribute Recognition,the key to Universal Neural Network

<p>To achieve the recognition of multi-attribute of object, I redesign the mnist dataset, change the color, size, location of the number. Meanwhile, I change the label accordingly. The deep neural network I use is the most common convolution neural network. Through test,we can conclude that we can use one neural network to recognize multi-attribute so long as the attribute difference of objects can be represented by functions. The concrete network(generation network) can generate the output which the input rarely contained from the attributes the network learned. Its generalization ability is good because the network is a continuous function. Through one more test, We can conclude that one neural network can do image recognition, speech recognition,and nature language processing and other things so long as the output node and the input node and more parameters add into the network. The network is universal so long as the network can process different inputs. I guess that the phenomenon of synaesthesia is the result of multi-input and multi-output. I guess that connection in mind can realize through the universal network and sending the output into input.<b></b></p>

Multi-attribute Recognition,the key to Universal Neural Network

<p>To achieve the recognition of multi-attribute of object, I redesign the mnist dataset, change the color, size, location of the number. Meanwhile, I change the label accordingly. The deep neural network I use is the most common convolution neural network. Through test,we can conclude that we can use one neural network to recognize multi-attribute so long as the attribute difference of objects can be represented by functions. The concrete network(generation network) can generate the output which the input rarely contained from the attributes the network learned. Its generalization ability is good because the network is a continuous function. Through one more test, We can conclude that one neural network can do image recognition, speech recognition,and nature language processing and other things so long as the output node and the input node and more parameters add into the network. The network is universal so long as the network can process different inputs. I guess that the phenomenon of synaesthesia is the result of multi-input and multi-output. I guess that connection in mind can realize through the universal network and sending the output into input.<b></b></p>

Multi-attribute Recognition,The key to Universal Neural Network

<p><b>To achieve the recognition of multi-attribute of object, I redesign the mnist dataset, change the color, size, location of the number. Meanwhile, I change the label accordingly.</b><b> </b><b>The deep neural network I use is the most common convolution neural network. Through test,we can conclude that we can use one neural network to recognize multi-attribute so long as the attribute difference of objects can be represented by functions. The </b><b>c</b><b>oncrete network(generation network) can generate the output which the input rarely contained from the attributes the network learned. Its generalization ability is good because the network is a continuous function. Through one more test, We can conclude that one neural network can do image recognition, speech recognition,and nature language processing and other things so long as the output node and the input node and more parameters add into the network. The network is universal so long as the network can process different inputs.</b><b> I guess that t</b><b>he phenomenon of synaesthesia is the result of multi-input and multi-output. </b><b>I guess that c</b><b>onnection in mind can realize through the universal network and sending the output into input.</b><b></b></p>

Sensor based Human Activity Recognition

Human activity recognition (HAR) is to recognize another person’s activities and it is one of the active research areas in the computer field. The goal of this System is to understand people's actions and interactions. We proposed a method of Human Activity is by predicting the person's activity, their personality, and their psychological state like Human activity recognition (HAR). We propose a recurrent neural network of deep learning architecture. The critical factor of RNN includes bidirectional connection that is simply called from the input node, the information only flows in forwarding direction after that it passthrough so many hidden layers to reach the output.. This system is to design the six different activities of a human. The final model should use as a good source of information about human's daily activities. The dataset has taken from UCI Machine Learning Repository. Our system accuracy is higher than the previous results.

Multi-attribute Recognition,The key to universal neural network

<p><b>To achieve the recognition of multi-attribute of object, I redesign the mnist dataset, change the color, size, location of the number. Meanwhile, I change the label accordingly.</b><b> </b><b>The deep neural network I use is the most common convolution neural network. Through test,we can conclude that we can use one neural network to recognize multi-attribute so long as the attribute difference of objects can be represented by functions. The </b><b>c</b><b>oncrete network(generation network) can generate the output which the input rarely contained from the attributes the network learned. Its generalization ability is good because the network is a continuous function. Through one more test, We can conclude that one neural network can do image recognition, speech recognition,and nature language processing and other things so long as the output node and the input node and more parameters add into the network. The network is universal so long as the network can process different inputs.</b><b> I guess that t</b><b>he phenomenon of synaesthesia is the result of multi-input and multi-output. </b><b>I guess that c</b><b>onnection in mind can realize through the universal network and sending the output into input.</b><b></b></p>

Multi-attribute Recognition,The key to universal neural network

<p><b>To achieve the recognition of multi-attribute of object, I redesign the mnist dataset, change the color, size, location of the number. Meanwhile, I change the label accordingly.</b><b> </b><b>The deep neural network I use is the most common convolution neural network. Through test,we can conclude that we can use one neural network to recognize multi-attribute so long as the attribute difference of objects can be represented by functions. The </b><b>c</b><b>oncrete network(generation network) can generate the output which the input rarely contained from the attributes the network learned. Its generalization ability is good because the network is a continuous function. Through one more test, We can conclude that one neural network can do image recognition, speech recognition,and nature language processing and other things so long as the output node and the input node and more parameters add into the network. The network is universal so long as the network can process different inputs.</b><b> I guess that t</b><b>he phenomenon of synaesthesia is the result of multi-input and multi-output. </b><b>I guess that c</b><b>onnection in mind can realize through the universal network and sending the output into input.</b><b></b></p>

Identifikasi Sinyal Elektromiografi Otot Vastus Medialis dan Erector Spinae dalam Transisi Gerakan untuk Kontrol Robot Kaki

Kehilangan beberapa bagian tubuh dan kelemahan otot akibat cedera adalah faktor yang mengganggu aktivitas manusia sehari-hari. Konsep exoskeleton adalah pendekatan yang sangat positif bagi manusia dalam hal kerusakan pada tungkai bawah. Dalam studi ini, ekstremitas bawah selama gerakan jongkok ke berdiri, berdiri ke duduk, duduk ke berdiri, dan berdiri ke jongkok menjadi fokus dalam penelitian ini. Sinyal elektromiografi terdeteksi dari vastus medialis dan erector spinae. Enam responden terlibat dalam melakukan percobaan ini. Ada 2 tahap dalam percobaan ini. Pada tahap pertama, gunakan fitur ekstraksi domain waktu seperti MAV, MAD, dan RMS. Latensi 500 ms dengan waktu tumpang tindih 10 ms digunakan. Ambang digunakan untuk mendeteksi awal kontraksi otot 0,002 mV dan bagian akhir kontraksi otot 0,0015 mV. Data dalam ambang batas digunakan sebagai input dari jaringan saraf tiruan. Penggunaan python 2.7 jaringan syaraf tiruan dibuat dengan 240 input node, 80 hidden node, dan 4 output node. Data pergerakan dengan total 556 digunakan untuk melatih jaringan. Data pergerakan dengan total 160 digunakan untuk menguji jaringan. Sistem ini mampu menginterpretasikan gerakan sebenarnya dengan nilai persentase 84% dan nilai kesalahan 16%. Pada tahap kedua menggunakan metode yang sama, sistem diuji dengan responden yang berbeda. Data pergerakan dengan total 104 digunakan untuk menguji jaringan. Persentase keberhasilan sistem dalam menafsirkan gerakan adalah 59% dan nilai kesalahan 41%.

Improvement of Mason Rule for Solving the Complicated Block Diagram

Background: This study first efficiently applies the previous result Chen Electrical Unifying Approach (CEUA) utilized in the basic circuit theory to construct the control system matrix equation of the complicated block diagram. Methods: Based on the simple matrix operations proposed in this study, we can easily derive the transfer function without using the traditional Mason rule and the reduced techniques of the block diagram. We have successfully proposed a unifying approach to improve the disadvantages of the Mason rule, in which all loops must be found out and only the transfer function between the input node and the output node is evaluated, and the shortcoming of the reduced techniques for the block diagram is that the calculating process is too complex to be accepted. Results: The salient features of the proposed method are that the transfer function of the complicated block diagram can be easily obtained without using traditional Mason rule and the transfer function of any two nodes is immediately derived within only one calculation. Conclusion: We compared some demonstrated examples with some traditional approaches. Moreover, to demonstrate the practical applicability, the study has investigated one practical example.

Design of Higher Order Delta Segma Modulator for ADC and CMOS Sensors

This paper presents a 4th-order incremental deltasigma ADC for CMOS image sensors. The ADC employing a cascade of integrators with feed forward (CIFF) architecture uses only one operational transconductance amplifier (OTA) by sharing the OTA between 1st and 2nd stages of the modulator. by using a proposed self-biasing amplifier ,which allows active signal summation at the quantizer input node without using an additional OTA, thus power and area savings are achieved. Fabricated in 90nm technology, the 4th orderdsm consumes 32.5 µW from a 1.2V supply.

Pengembangan Model Jaringan Syaraf Tiruan untuk Mendeteksi Anomali Satelit LAPAN-TUBSAT

Kasus satelit mengalami anomali seringkali di temukan pada satelit-satelit yang beroperasi pada orbit polar. Namun permasalahan yang muncul adalah kondisi satelit sering berubah-ubah sehingga operator belum bisa mengantisipasi kondisi tersebut. Oleh sebab itu, model deteksi kondisi satelit dapat berperan sebagai early warning operator satelit untuk mempersiapkan strategi yang berkaitan dengan kebijakan preventif terkait pencegahan ketika satelit mengalami kondisi tidak normal. Tujuan penelitian ini adalah menerapkan Jaringan Syaraf Tiruan (JST) bakcpropagation dalam mendeteksi kondisi anomali pada satelit LAPAN-TUBSAT, serta mengetahui tingkat akurasi dari proses deteksi tersebut sehingga diperoleh parameter dan arsitektur jaringan JST terbaik. Proses pembelajaran dan pengujian JST menggunakan data kejadian anomali tahun 2009 sampai 2014. Arsitektur JST yang digunakan adalah jumlah node input 4, dua hidden layer, jumlah node lapisan tersembunyi (hidden neuron) divariasikan pada nilai 5, 10, 15 dan 20. Parameter yang diberikan pada proses pembelajaran antara lain adalah fungsi aktivasi, toleransi galat, jumlah epoch maksimal dan variasi nilai laju pembelajaran (learning rate). Empat parameter input yang digunakan yakni elektron (mep0e1), proton (mep0p1), indeks Kp serta indeks Dst. Hasil penelitian menunjukkan bahwa arsitektur jaringan syaraf terbaik dihasilkan oleh jaringan dengan jumlah input node empat, hidden neuron 20 dan 10, nilai learning rate sebesar 0.05 dengan 306 epoch, menghasilan rata-rata akurasi sebesar 98.13%, serta nilai precision dan recall sebesar 98.21% dan 94.81%.