365 days of Artificial Intelligence and Machine Learning
This is the 365 days Challenge of Machine Learning, Deep Learning, AI, and Optimization (mini-projects and research papers) that I picked up at the start of January 2022. I have used various environments and Google Colab, and certain environments for this work as it required various libraries and datasets to be downloaded. The following are the problems that I tackled:
Tug-Of-War Optimization (Kaveh, A., & Zolghadr, A. (2016). A novel meta-heuristic algorithm: tug of war optimization. Iran University of Science & Technology, 6(4), 469-492.)
Nuclear Reaction Optimization (Wei, Z., Huang, C., Wang, X., Han, T., & Li, Y. (2019). Nuclear Reaction Optimization: A novel and powerful physics-based algorithm for global optimization. IEEE Access.)
+ So many equations and loops - take time to run on larger dimension
+ General O (g * n * d)
+ Good convergence curse because the used of gaussian-distribution and levy-flight trajectory
+ Use the variant of Differential Evolution
Henry Gas Solubility Optimization (Hashim, F. A., Houssein, E. H., Mabrouk, M. S., Al-Atabany, W., & Mirjalili, S. (2019). Henry gas solubility optimization: A novel physics-based algorithm. Future Generation Computer Systems, 101, 646-667.)
+ Too much constants and variables
+ Still have some unclear point in Eq. 9 and Algorithm. 1
+ Can improve this algorithm by opposition-based and levy-flight
+ A wrong logic code in line 91 "j = id % self.n_elements" => to "j = id % self.n_clusters" can make algorithm converge faster. I don't know why?
+ Good results come from CEC 2014
Queuing Search Algorithm (Zhang, J., Xiao, M., Gao, L., & Pan, Q. (2018). Queuing search algorithm: A novel metaheuristic algorithm for solving engineering optimization problems. Applied Mathematical Modelling, 63, 464-490.)
Day 16 (01/16/2022): Evolutionary Optimization algorithms
Explored the contents of Human Activity-based optimization techniques such as:
Genetic Algorithms (Holland, J. H. (1992). Genetic algorithms. Scientific american, 267(1), 66-73)
Differential Evolution (Storn, R., & Price, K. (1997). Differential evolution–a simple and efficient heuristic for global optimization over continuous spaces. Journal of global optimization, 11(4), 341-359)
Coral Reefs Optimization Algorithm (Salcedo-Sanz, S., Del Ser, J., Landa-Torres, I., Gil-López, S., & Portilla-Figueras, J. A. (2014). The coral reefs optimization algorithm: a novel metaheuristic for efficiently solving optimization problems. The Scientific World Journal, 2014)
Particle Swarm Optimization (Eberhart, R., & Kennedy, J. (1995, October). A new optimizer using particle swarm theory. In MHS'95. Proceedings of the Sixth International Symposium on Micro Machine and Human Science (pp. 39-43). IEEE)
Cat Swarm Optimization (Chu, S. C., Tsai, P. W., & Pan, J. S. (2006, August). Cat swarm optimization. In Pacific Rim international conference on artificial intelligence (pp. 854-858). Springer, Berlin, Heidelberg)
Whale Optimization (Mirjalili, S., & Lewis, A. (2016). The whale optimization algorithm. Advances in engineering software, 95, 51-67)
Bacterial Foraging Optimization (Passino, K. M. (2002). Biomimicry of bacterial foraging for distributed optimization and control. IEEE control systems magazine, 22(3), 52-67)
Adaptive Bacterial Foraging Optimization (Yan, X., Zhu, Y., Zhang, H., Chen, H., & Niu, B. (2012). An adaptive bacterial foraging optimization algorithm with lifecycle and social learning. Discrete Dynamics in Nature and Society, 2012)
Artificial Bee Colony (Karaboga, D., & Basturk, B. (2007, June). Artificial bee colony (ABC) optimization algorithm for solving constrained optimization problems. In International fuzzy systems association world congress (pp. 789-798). Springer, Berlin, Heidelberg)
Pathfinder Algorithm (Yapici, H., & Cetinkaya, N. (2019). A new meta-heuristic optimizer: Pathfinder algorithm. Applied Soft Computing, 78, 545-568)
Harris Hawks Optimization (Heidari, A. A., Mirjalili, S., Faris, H., Aljarah, I., Mafarja, M., & Chen, H. (2019). Harris hawks optimization: Algorithm and applications. Future Generation Computer Systems, 97, 849-872)
Sailfish Optimizer (Shadravan, S., Naji, H. R., & Bardsiri, V. K. (2019). The Sailfish Optimizer: A novel nature-inspired metaheuristic algorithm for solving constrained engineering optimization problems. Engineering Applications of Artificial Intelligence, 80, 20-34)
Credits (from Day 14--17): Learnt a lot due to Nguyen Van Thieu and his repository that deals with metaheuristic algorithms. Plan to use these algorithms in the problems enountered later onwards.
Day 44 (02/13/2022): Image Deraining Implementation using SPANet
Referred from: RESCAN by Xia Li et al. The CUDA extension references pyinn by Sergey Zagoruyko and DSC(CF-Caffe) by Xiaowei Hu!!
Day 47 (02/16/2022): img2pose: Face Alignment and Detection via 6DoF, Face Pose Estimation This repository draws directly from the one mentioned here. I've tried implementing it on different datasets such as the BIWI ad AWFL dataset. Furthermore, the models weren't trained from scratch. The run was meant to be a way to report the numbers in the paper.
Paper accepted to the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2021
Figure 1: We estimate the 6DoF rigid transformation of a 3D face (rendered in silver), aligning it with even the tiniest faces, without face detection or facial landmark localization. Our estimated 3D face locations are rendered by descending distances from the camera, for coherent visualization.
Summary: This repository provides a novel method for six degrees of fredoom (6DoF) detection on multiple faces without the need of prior face detection. After prediction, one can visualize the detections (as show in the figure above), customize projected bounding boxes, or crop and align each face for further processing. See details below.
We propose real-time, six degrees of freedom (6DoF), 3D face pose estimation without face detection or landmark localization. We observe that estimating the 6DoF rigid transformation of a face is a simpler problem than facial landmark detection, often used for 3D face alignment. In addition, 6DoF offers more information than face bounding box labels. We leverage these observations to make multiple contributions: (a) We describe an easily trained, efficient, Faster R-CNN--based model which regresses 6DoF pose for all faces in the photo, without preliminary face detection. (b) We explain how pose is converted and kept consistent between the input photo and arbitrary crops created while training and evaluating our model. (c) Finally, we show how face poses can replace detection bounding box training labels. Tests on AFLW2000-3D and BIWI show that our method runs at real-time and outperforms state of the art (SotA) face pose estimators. Remarkably, our method also surpasses SotA models of comparable complexity on the WIDER FACE detection benchmark, despite not been optimized on bounding box labels.
请发表评论