1. Globoki modeli za detekcijo in prepoznavo obrazov v video vsebinah in slikah : magistrsko deloStefani Bojanić, 2025, master's thesis Abstract: Človeški obraz predstavlja eno od najpomembnejših biometričnih značilnosti, saj združuje informacijo o identiteti, spolu, starosti in čustvenem izrazu. V tem okviru se detekcija in prepoznavanje obrazov kažeta kot dva neločljivo povezana procesa. V magistrski nalogi so predstavljeni ključni izzivi tega področja ter stanje razvoja, ki zajema vse od klasičnih metod do sodobnih pristopov z globokim učenjem, s poudarkom na konvolucijskih nevronskih mrežah. Razvoj in eksperimenti so bili izvedeni s programskim jezikom Python. V okolju Visual Studio Code smo tako razvili sistem za prepoznavanje obrazov z uporabo algoritma ArcFace.
Keywords: detekcija obrazov, prepoznavanje obrazov, globoko učenje, state of the art, konvolucijske nevronske mreže (CNN), ArcFace
Published in DKUM: 23.10.2025; Views: 0; Downloads: 12
 Full text (4,66 MB) |
2. Robot for navigation in maize crops for the Field Robot Event 2023David Iván Sánchez-Chávez, Noé Velázquez-López, Guillermo García-Sánchez, Alan Hernández-Mercado, Omar Alexis Avendaño-Lopez, Mónica Elizabeth Berrocal-Aguilar, 2024, original scientific article Abstract: Navigation in a maize crop is a crucial task for the development of autonomous robots in agriculture, with numerous applications such as spraying, monitoring plant growth and health, and detecting weeds and pests. The Field Robot Event 2023 (FRE) continued to challenge universities and other research teams to push the development of algorithms for agricultural robots further. The Universidad Autónoma Chapingo has been developing a robot for various agricultural tasks, aiming to provide a low-cost alternative to work with Mexican farmers in the future. For this edition of the FRE, a navigation algorithm was created using an encoder, an IMU (Inertial Measurement Unit), an RPLIDAR (Rotating Platform Light Detection and Ranging), and cameras to collect data for decision-making. The algorithm was developed in ROS Melodic, dividing the task into steps that were tested to determine the robot's actual movements. The system navigates by using ROIs (regions of interest) and the mass center to guide the robot between maize rows. It calculates the mean of the final orientation values before reaching the end of a row, which is detected using an RPLIDAR. For turns and straight-line movements to reach the next row, the orientation is used as a guide. To detect plants for spraying, lasers located on each side of the vehicle are employed. Obstacle detection relies on a YOLOv5 (You Only Look Once) trained model and a laser, while reverse navigation uses a rear camera. During the competition, the robot faced challenges such as dealing with grass, the small size of the plants, and the need to use a different power source, which affected its performance.
Keywords: machine vision, convolutional neural network (CNN), regions of interest (ROI), autonomous navigation
Published in DKUM: 23.04.2025; Views: 0; Downloads: 3
 Link to file |
3. CNN-Based Vessel Meeting Knowledge Discovery From AIS Vessel TrajectoriesPeng Chen, Shuang Liu, Niko Lukač, 2023, original scientific article Abstract: How to extract a collection of trajectories for different vessels from the raw AIS data to discover vessel meeting knowledge is a heavily studied focus. Here, the AIS database is created based on the raw AIS data after parsing, noise reduction and dynamic Ramer-Douglas-Peucker compression. Potential encountering trajectory pairs will be recorded based on the candidate meeting vessel searching algorithm. To ensure consistent features extracted from the trajectories in the same time period, time alignment is also adopted. With statistical analysis of vessel trajectories, sailing segment labels will be added to the input feature. All motion features and sailing segment labels are combined as input to one trajectory similarity matching method based on convolutional neural network to recognize crossing, overtaking or head-on situations for each potential encountering vessel pair, which may lead to collision if false actions are adopted. Experiments on AIS data show that our method is effective in classifying vessel encounter situations to provide decision support for collision avoidance.
Keywords: AIS Data, CNN, Dynamic Rammer-Douglas-Peucker, knowledge discovery, maneuvering pattern, traffic pattern, trajectory
Published in DKUM: 19.03.2024; Views: 538; Downloads: 421
Full text (3,84 MB)
This document has many files! More... |
4. Napovedovanje prodaje zdravil z uporabo naprednih metod napovedovanja časovnih vrst : magistrsko deloŽan Pudič, 2023, master's thesis Abstract: Magistrsko delo zajema uporabo naprednih modelov za napovedovanje prodaje zdravil. Cilj dela je s pomočjo naprednih metod napovedovanja v programskem okolju R, postaviti napovedovalne modele za posamezne skupine zdravil, ki bodo v izbranih intervalih zaupanja uspešno napovedali prihodnjo prodajo. V delu smo za potrebe napovedovanja uporabili modele kot so ARIMA, CNN, Holt-Winters pri čemer smo te primerjali z naivno metodo napovedovanj in tako ocenili njihovo sposobnost napovedovanja. Prav tako smo v delu podrobno analizirali izhode v fazi kreiranja modelov na podlagi katerih smo izvedli nadaljnjo selekcijo modelov s katerimi lahko uspešno napovemo prihodnjo prodajo. Uspešne napovedi smo izvedli pri vseh skupinah zdravil. V delu je najuspešnejšo napoved pri skupinah zdravil M01AB. M01AE, N02BA, N05C, R03 in R06 imel ARIMA model. Prodajo v skupinah zdravil N05C in N02BE pa lahko napovedujemo zgolj z uporabo CNN, saj noben izmed preostalih modelov ni uspel dovolj dobro zajeti informacij znotraj časovne vrste, da bi lahko z njim podali napoved.
Keywords: napredne metode napovedovanja, časovne vrste, napovedovanje prodaje, napovedovalni modeli, ARIMA, CNN, Holt-Winters.
Published in DKUM: 26.09.2023; Views: 436; Downloads: 92
Full text (3,58 MB) |
5. Zaznava in lociranje malin z uporabo YOLO algoritma : magistrsko deloUrban Kenda, 2023, master's thesis Abstract: V magistrskem delu smo raziskali delovanje LiDAR senzorjev ter uporabo umetne inteligence v strojnem vidu, vključno z nevronskimi mrežami, konvolucijskimi nevronskimi mrežami (CNN) in algoritmi YOLOv3, v4 in v4-tiny. V praktičnem delu smo testirali vse tri algoritme in nato izbrali najuspešnejšega, YOLOv4, ter ga dodatno analizirali. Preverili smo hitrost algoritmov ter razvili algoritem, ki je na podlagi oblakov točk in kamere sposoben določiti lokacijo malin. Ugotovili smo, da je uporaba LiDAR senzorjev v kombinaciji z umetno inteligenco učinkovita pri zaznavanju in lociranju malin v 3D-prostoru. Najuspešnejši algoritem YOLOv4 je bil sposoben razvrstiti zrele in nezrele maline z natančnostjo 84,13 %. Naš razviti algoritem je omogočil določanje lokacije malin s kombinirano uporabo oblakov točk in kamere ter tako skoraj v polovici izmerjenih primerov določil lokacijo z napako, manjšo od 2 cm.
Keywords: malina, strojni vid, YOLO, nevronska mreža, CNN, oblak točk
Published in DKUM: 15.06.2023; Views: 512; Downloads: 126
Full text (1013,36 KB) |
6. Sortiranje fižolovih zrn z uporabo strojnega vida : magistrsko deloSamo Šlander, 2022, master's thesis Abstract: V magistrskem delu je predstavljen projekt, ki temelji na zasnovi metode za sortiranje fižola s pomočjo strojnega vida. Njegov namen je nadomestiti ročno sortiranje. V delu je opisan tudi razvoj klasifikacijskega modela za razvrščanje slik fižola ter mehanskega koncepta. Prav tako je predstavljeno področje strojnega vida ter strojnega učenja. Na koncu pa je predstavljen nabor baze podatkov, razvrstitev slik po kategorijah z uporabo prenosnega učenja in algoritma k-najbližjih sosedov, ovrednotenje rezultatov le-teh ter koncept mehanske separacije.
Keywords: strojni vid, CNN, KNN, sortiranje fižola, klasifikacija
Published in DKUM: 28.06.2022; Views: 874; Downloads: 203
Full text (5,66 MB) |
7. Development of a Model for Predicting Brake Torque Using LSTM and TCN Models : magistrsko deloTomaž Roškar, 2020, master's thesis Abstract: The main purpose of this thesis is to compare two state-of-the-art machine learning models, LSTM (Long Short-Term Memory) and TCN (Temporal Convolutional Network), on an AVL List GmbH case use, where the goal is to predict vehicle brake torque. Dataset used for model testing consists of multiple features which are preprocessed using several preprocessing methods. For model implementation Python’s libraries Keras and TensorFlow are used. Results from this thesis show that TCN is able to outperform LSTM. TCN achieves lower RMSE on the test dataset and is significantly faster in training and evaluation.
Keywords: brake torque, machine learning, neural network, LSTM, TCN, RNN, CNN
Published in DKUM: 24.09.2020; Views: 1392; Downloads: 14
Full text (3,33 MB) |
