Introduction In this experiment, we will be using VGG19 which is pre-trained on ImageNet on Cifar-10 dataset. We will be using PyTorch for this experiment. (A Keras version is also available) VGG19 is well known in producing promising results due to the depth of it. The “19” comes from the number of layers it has. […]Read more "Transfer Learning of VGG19 on Cifar-10 Dataset using PyTorch"
Introduction In this Lab, we will be implementing Network In Network  where its purpose is to enhance model discriminability for local patches within the receptive field. Conventional convolutional layers uses linear filters followed by a nonlinear activation function. The downside of the conventional method is the local receptors are too simple and doesn’t project local […]Read more "Network-in-Network Implementation using TensorFlow"
Wiener filter is a filter used to remove degradation from a image without being affected by noise that’s present in the image. If wiener filter is not used, and the image is restored just by dividing the frequency domain of an image by the frequency domain of the future, the image will be badly corrupted […]Read more "Wiener Filter"
In this discussion, we’re trying to reduce noise in the image and in the same time sharpen our image. The image that we’ll be using will look this: As we can see, some important details of the image is actually dark. What we want to do is first make the shady portion of the image […]Read more "Noise Cancellation and Sharpening"
Introduction White balance is used to remove unrealistic colour casts caused by different light source. It’s easy for human eyes to adjust what we see according to the light source of the current environment but it’s hard for the camera to adjust according to the lighting condition of the room. Sometimes, we refer that as […]Read more "Image White Balance"
Introduction There’re a few ways to analyze an image. Few of them are CIE LUV, CIE LAB and converting them to HSI, HSV, HSL, etc. CIE LUV and CIE LAB are similar for the case of “L” which is responsible for Lightness of an image. A and B (U and V) are responsible for red/green opponent […]Read more "Image Transformation based on HSI"
Each pixel of an image is usually representing using 4 bytes: 3 colors (R,G,B) and Alpha (A) or just 3 bytes (R,G,B). As we know 1 byte is 8 bits and it can represent values from 0 to 255 (unsigned). Thus in terms of RGB a pixel will have a combination of 224 which is 16777216 […]Read more "Resolution Quantization"