Spatial Based Deep Learning Autonomous  Wheel Robot Using CNN

Prasetyo, Eko; Nambo, Hidetaka; Prasetya, Dwi Arman; Dirgantara, Wahyu; Windi, Hari Fitria

doi:10.24843/lkjiti.2020.v11.i03.p05

Cited by 7 publications

(9 citation statements)

References 17 publications

(22 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Inference uses the trained model to detect objects in new images or data by generating predictions about the location and class of things in the image or data [26]. As shown in Figure 6, the lettuce inference aims to identify lettuce leaves in the image and provide a bounding box for each detected leaf.…”

Section: Inference Object Detectionmentioning

confidence: 99%

A Fine-Tuned RetinaNet for Real-Time Lettuce Detection

Prasetyo,

Nambo

2024

LKJITI

View full text Add to dashboard Cite

The agricultural industry plays a vital role in the global demand for food production. Along with population growth, there is an increasing need for efficient farming practices that can maximize crop yields. Conventional methods of harvesting lettuce often rely on manual labor, which can be time-consuming, labor-intensive, and prone to human error. These challenges lead to research into automation technology, such as robotics, to improve harvest efficiency and reduce reliance on human intervention. Deep learning-based object detection models have shown impressive success in various computer vision tasks, such as object recognition. RetinaNet model can be trained to identify and localize lettuce accurately. However, the pre-trained models must be fine-tuned to adapt to the specific characteristics of lettuce, such as shape, size, and occlusion, to deploy object recognition models in real-world agricultural scenarios. Fine-tuning the models using lettuce-specific datasets can improve their accuracy and robustness for detecting and localizing lettuce. The data acquired for RetinaNet has the highest accuracy of 0.782, recall of 0.844, f1-score of 0.875, and mAP of 0,962. Metrics evaluate that the higher the score, the better the model performs.

show abstract

Section: Inference Object Detectionmentioning

confidence: 99%

A Fine-Tuned RetinaNet for Real-Time Lettuce Detection

Prasetyo,

Nambo

2024

LKJITI

View full text Add to dashboard Cite

show abstract

“…Convolutional Neural Network (CNN) is a DNN architecture used to process data in the form of a grid [33]. CNN consists of convolution layers whose function is to map the characteristics of the input data through it.…”

Section: Deep Learningmentioning

confidence: 99%

Analysis of Public Sentiment Towards Goverment Efforts to Break the Chain of Covid-19 Transmission in Indonesia Using CNN and Bidirectional LSTM

Jaluwana

Sasmita

Suarjaya

2022

J. RESTI (Rekayasa Sist. Teknol. Inf.)

View full text Add to dashboard Cite

COVID-19 is a new disease that has a negatively impacts in Indonesia, so the government is taking several measures to suppress the spread of COVID-19, such as new normal, social distancing, health protocols fines, and COVID-19 vaccination. The government's handling efforts have reaped a variety of negative to positive responses from the public on social media, so this study aims to determine the effectiveness of the government's efforts by analyzing public sentiment using the Deep Learning method with 1,875 training datasets consisting of four types government efforts and taken from various media social. The use of Deep Learning begins with testing several Deep Learning architectures to determine the best architecture for predicting data. The architectures tested include CNN and Bi-LSTM, where from these tests, Bi-LSTM outperforms CNN with the best performance achieving the accuracy of 97.34% and 97.33% for precision, recall, and F1-score. The results of public sentiment analysis show that social distancing efforts are considered the most effective by obtaining the most positive sentiments by 33.93%, while the effort to health protocol fines is considered lacking because it obtains the most negative sentiment of 35.64%, so the government must continue to enforce social distancing and optimize other efforts that are still considered ineffective.

show abstract

“…[14] Secara umum algoritma backpropagation standar menggunakan fungsi pelatihan gradient discent, yaitu : traingda, traingdm, traingd , trainngdx, [15] dan banyak lagi fungsi pelatihan yang dapat digunakan dalam optimasi yang mempengaruhi hasil komputasi, misalnya gradien konjugasi berskala (trainscg), [16] fungsi pelatihan urutan acak dengan fungsi pembelajaran (trainr), [17] regulasi bayesian (trainbr), [18] pelatihan bias perintah acak yang tak terawasi (trainru), [19] levenberg marquardt (trainlm), [20], [21] one step secant (trainos), [22] fungsi pembelajaran inkremental berurutan (trains), [23] fungsi pelatihan tanpa pengawasan (trainbu), fungsi batch memakai aturan pembelajaran bias (trainb), [24] fungsi konjugasi gradien (traincgf, traincgb, traingcgp), [25] resilient (trainrp), [26] BFGS quasi newton, [27] alur siklus bias (trainc), [28] BFGS quasi newton referensi adaftif kontrol (trainbfgc). [29], [30] Menggunakan fungsi pelatihan dan transfer menghasilkan tingkat keakuratan peramalan yang berbeda-beda dengan menggunakan parameter atau metode, dan data yang digunakan dalam pengujian. [31]- [34] Makalah ini fokus dalam membahas penggunaan fungsi pelatihan algoritma Backpropagation dalam memprediksi prestasi mahasiswa.…”

Section: Pendahuluanunclassified

Prediksi Prestasi Mahasiswa Dengan Menggunakan Algoritma Backpropagation

Sonang¹,

Purba²,

Sirait³

2022

Tekinkom

View full text Add to dashboard Cite

This study aims to overcome the problems in predicting student achievement at the Polytechnic Business Indonesia Pematangsiantar. To predict student achievement is done by applying Backpropagation algorithm and implement it into Matlab software. Backpropagation algorithm is one of the methods on artificial neural networks that is quite reliable in solving problems including prediction. In this study conducted on the object of students semester One with a lot of data samples 26 samples. The data sample is divided into two parts, 70% of the data is used as training data and 30% of the data is used as testing data. This study uses ten architectural models, namely 9-2-1, 9-3-1, 9-4-1, 9-5-1, 9-6-1, 9-7-1, 9-8-1, 9-9-1, 9-10-1, 9-11-1. Of the ten Backpropagation network architecture models implemented in predicting student achievement in Matlab software obtained the best output is 9-2-1 pattern with epoch 8149, time duration for 17 seconds, and MSE (error rate) value of 2.80 e-05 for training and MSE (error rate) of 0.1248 with accuracy of 87.5% for testing. The best architecture obtained is expected to be used as a picture by the academic Polytechnic Business Indonesia (PBI) in predicting student achievement.

show abstract

Spatial Based Deep Learning Autonomous Wheel Robot Using CNN

Cited by 7 publications

References 17 publications

A Fine-Tuned RetinaNet for Real-Time Lettuce Detection

A Fine-Tuned RetinaNet for Real-Time Lettuce Detection

Analysis of Public Sentiment Towards Goverment Efforts to Break the Chain of Covid-19 Transmission in Indonesia Using CNN and Bidirectional LSTM

Prediksi Prestasi Mahasiswa Dengan Menggunakan Algoritma Backpropagation

Contact Info

Product

Resources

About