Search Results for author:
Found 7 papers, 3 papers with code
Frame by Familiar Frame: Understanding Replication in Video Diffusion Models
In our paper, we present a systematic investigation into the phenomenon of sample replication in video diffusion models.
Analyzing Bias in Diffusion-based Face Generation Models
Diffusion models are becoming increasingly popular in synthetic data generation and image editing applications.
SAR Despeckling using a Denoising Diffusion Probabilistic Model
The despeckled image is recovered by a reverse process which iteratively predicts the added noise using a noise predictor which is conditioned on the speckled image.
SAR Despeckling Using Overcomplete Convolutional Networks
We show that the proposed network improves despeckling performance compared to recent despeckling methods on synthetic and real SAR images.
Transformer-based SAR Image Despeckling
Synthetic Aperture Radar (SAR) images are usually degraded by a multiplicative noise known as speckle which makes processing and interpretation of SAR images difficult.
A Thickness Sensitive Vessel Extraction Framework for Retinal and Conjunctival Vascular Tortuosity Analysis
In addition, the proposed framework was utilized to determine the association of diabetes with retinal and conjunctival vascular tortuosity.
A Joint Convolutional and Spatial Quad-Directional LSTM Network for Phase Unwrapping
Phase unwrapping is a classical ill-posed problem which aims to recover the true phase from wrapped phase.
