I implemented Latent Diffusion Models (LDM) capable of both image-to-image and text-to-image generation. Additionally, I utilized various schedulers and integrated in-model techniques (such as DDPM, DDIM). Moreover, I implemented a U-Net architecture integrated with Cross-Attention layers to enable

We are replicating the model described in the paper and improving model accuracy by exploring various network architectures, such as Spatial Pyramid Pooling, Atrous Convolutions, and Encoder-Decoder structures.

Using convolutional neural network(CNN) PyTorch, I classified images of different clothes on 10 different image classes.

Binary classification of given data using deep neural network and ReLU non-linearity. 98% accuracy

I finished the Supervised Machine Learning: Classification and Regression, course from coursera successfully