|| Image denoising to estimate the original signal is a well-studied problem in consumer electronics. The recent studies in the field are focusing on the benchmarking denoising techniques on the datasets which closely reproduce the real ground truth and noise conditions, instead of using the traditional way of evaluating on the artificially added noise. In contrast, the image denoising for Transmission Electron Microscopy (TEM) lacks far behind. We believe the two main reason behind this is the unavailability of noise-free ground truth and the standardized publicly available dataset. In an attempt to address these limitations, we are releasing a standardized dataset having cilia cross-section instances extracted from high-resolution TEM images. We also define a strategy to generate the synthetic noise-free images utilizing multiple instances. Furthermore, we presented a benchmark comparative study using classical denoising filters to perform edge-preserving denoising of cilia cross-sections.