Source author record

Andry Rasoanaivo

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Discrete Mathematics eess.IV Machine Learning math.CO

Catalog footprint

Research overview

2works

4topics

3close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2020arXiv

ESRGAN+ : Further Improving Enhanced Super-Resolution Generative Adversarial Network

Enhanced Super-Resolution Generative Adversarial Network (ESRGAN) is a perceptual-driven approach for single image super resolution that is able to produce photorealistic images. Despite the visual quality of these generated images, there is still room for improvement. In this fashion, the model is extended to further improve the perceptual quality of the images. We have designed a novel block to replace the one used by the original ESRGAN. Moreover, we introduce noise inputs to the generator network in order to exploit stochastic variation. The resulting images present more realistic textures. The code is available at https://github.com/ncarraz/ESRGANplus .

preprint2016arXiv

The Maximum Block Size of Critical Random Graphs

Let $G(n,\, M)$ be the uniform random graph with $n$ vertices and $M$ edges. Let $B_n$ be the maximum block-size of $G(n,\, M)$ or the maximum size of its maximal $2$ -connected induced subgraphs. We determine the expectation of $B_n$ near the critical point $M=n/2$ . As $n-2M \gg n^{2/3}$ , we find a constant $c_1$ such that $c_1 = \lim_{n \rightarrow \infty} \left(1 - \frac{2M}{n} \right) \, E B_n \, .$ Inside the window of transition of $G(n,\, M)$ with $M=\frac{n}{2}(1+λn^{-1/3})$ , where $λ$ is any real number, we find an exact analytic expression for $c_2(λ) = \lim_{n \rightarrow \infty} \frac{E B_n} {n^{1/3}} \, .$ This study relies on the symbolic method and analytic tools coming from generating function theory which enable us to describe the evolution of $n^{-1/3} \, E B_n$ as a function of $λ$ .

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.