文件列表:

pa1_NCC_Image_Layering (0, 2023-01-28)
pa1_NCC_Image_Layering\.ipynb_checkpoints (0, 2023-01-28)
pa1_NCC_Image_Layering\.ipynb_checkpoints\pa1-checkpoint.ipynb (7175480, 2023-01-28)
pa1_NCC_Image_Layering\data (0, 2023-01-28)
pa1_NCC_Image_Layering\data\.DS_Store (6148, 2023-01-28)
pa1_NCC_Image_Layering\data\00125v.jpg (55764, 2023-01-28)
pa1_NCC_Image_Layering\data\00149v.jpg (71320, 2023-01-28)
pa1_NCC_Image_Layering\data\00153v.jpg (59455, 2023-01-28)
pa1_NCC_Image_Layering\data\00351v.jpg (77596, 2023-01-28)
pa1_NCC_Image_Layering\data\00398v.jpg (64600, 2023-01-28)
pa1_NCC_Image_Layering\data\01112v.jpg (60624, 2023-01-28)
pa1_NCC_Image_Layering\data\emir.tif (71288266, 2023-01-28)
pa1_NCC_Image_Layering\meta_data (0, 2023-01-28)
pa1_NCC_Image_Layering\meta_data\example-Prokudin-Gorskii.png (134549, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid (0, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid\00125v.jpg (47137, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid\00149v.jpg (56240, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid\00153v.jpg (46458, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid\00351v.jpg (62326, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid\00398v.jpg (52891, 2023-01-28)
pa1_NCC_Image_Layering\output-pyramid\01112v.jpg (45606, 2023-01-28)
pa1_NCC_Image_Layering\output (0, 2023-01-28)
pa1_NCC_Image_Layering\output\00125v.jpg (46291, 2023-01-28)
pa1_NCC_Image_Layering\output\00149v.jpg (54914, 2023-01-28)
pa1_NCC_Image_Layering\output\00153v.jpg (45355, 2023-01-28)
pa1_NCC_Image_Layering\output\00351v.jpg (61104, 2023-01-28)
pa1_NCC_Image_Layering\output\00398v.jpg (51606, 2023-01-28)
pa1_NCC_Image_Layering\output\01112v.jpg (45196, 2023-01-28)
pa1_NCC_Image_Layering\pa1.ipynb (7175480, 2023-01-28)
pa2_SIFT_Feature_Matching (0, 2023-01-28)
pa2_SIFT_Feature_Matching\data (0, 2023-01-28)
pa2_SIFT_Feature_Matching\data\1a_notredame.jpg (948494, 2023-01-28)
pa2_SIFT_Feature_Matching\data\1b_notredame.jpg (708480, 2023-01-28)
pa2_SIFT_Feature_Matching\data\2a_rushmore.jpg (1619620, 2023-01-28)
pa2_SIFT_Feature_Matching\data\2b_rushmore.jpg (1745464, 2023-01-28)
pa2_SIFT_Feature_Matching\data\3a_gaudi.jpg (1731934, 2023-01-28)
pa2_SIFT_Feature_Matching\data\3b_gaudi.jpg (317174, 2023-01-28)
... ...

# Computer Vision This repo holds all programming assignments completed for my Computer Vision course (Fall 2021). ***Note:*** Scaffolding code was given for some of these assignments. All code beneath functions and methods is completed by me, unless otherwise noted. # Assignment Descriptions ## PA1 --- NCC Image Layering Layering of red, green and blue image filters based on a normalized cross correlation (NCC) calculation of image pyramids. - **Code:** [`pa1_NCC_Image_Layering/pa1.ipynb`](https://github.com/luke-davidson/ComputerVision/blob/main/pa1_NCC_Image_Layering/pa1.ipynb) ## PA2 --- SIFT Feature Extraction + Matching Implementation of SIFT feature extraction and corner matching: sobel kernels, second moment matrices, max pooling, patch descriptors, pairwise feature distances, gradient magnitudes and orientations, feature matching. - **Code:** [`pa2_SIFT_Feature_Matching/pa2_code/pa2.ipynb`](https://github.com/luke-davidson/ComputerVision/blob/main/pa2_SIFT_Feature_Matching/pa2_code/pa2.ipynb) - **Report:** [`pa2_SIFT_Feature_Matching/pa2_report.pdf`](https://github.com/luke-davidson/ComputerVision/blob/main/pa2_SIFT_Feature_Matching/pa2_report.pdf) ## PA3 --- Camera Calibration + Fundamental Matrix Estimation Estimating camera projection and fundamental matrices using RANSAC to match images at different viewpoints. - **Code:** [`pa3_CameraCal_RANSAC_Matrix_Estimations/pa3_code/pa3.ipynb`](https://github.com/luke-davidson/ComputerVision/blob/main/pa3_CameraCal_RANSAC_Matrix_Estimations/pa3_code/pa3.ipynb) - **Report:** [`pa3_CameraCal_RANSAC_Matrix_Estimations/pa3_report.pdf`](https://github.com/luke-davidson/ComputerVision/blob/main/pa3_CameraCal_RANSAC_Matrix_Estimations/pa3_report.pdf) ## PA4 --- Fully Connected Neural Network (FCN) Linear Classifier: Implementation of a linear classifier using cross entropy and SVM Hinge losses. FC Net: Implementation of a fully connected neural network. Processes implemented include forward and backward passes of affine layers, ReLU activation, softmax loss, SGD and momentum. - **Code:** - [`pa4_Fully_Connected_Net/pa4_code/linear_classifier.ipynb`](https://github.com/luke-davidson/ComputerVision/blob/main/pa4_Fully_Connected_Net/pa4_code/linear_classifier.ipynb) - [`pa4_Fully_Connected_Net/pa4_code/fc_net.ipynb`](https://github.com/luke-davidson/ComputerVision/blob/main/pa4_Fully_Connected_Net/pa4_code/fc_net.ipynb) ## PA5 --- Convolutional Neural Network (CNN) Full naive implementation of a CNN. Naive implementations of forward and backward passes of convolution layers, batch normalization (normal and spatial) and adaptive average pooling. Network trained using cross-validation. - **Code:** [`pa5_CNN/pa5_1_code/cnn.ipynb`](https://github.com/luke-davidson/ComputerVision/blob/main/pa5_CNN/pa5_1_code/cnn.ipynb)

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：