Generating Stylistic Images by Extending Neural Style Transfer Method
Bibliographic record
Abstract
Fine arts have long been considered a reserved mastery for the minority of talented individuals in society. The ability to create paintings using unique visual components such as color, stroke, theme, and other creative aspects is currently beyond the reach of computer algorithms. However, there exist algorithms, which have the capability of imitating an artist's painting style and stamping it on to virtually any image to create a one-of-a-kind piece. This paper introduces the concept of using a convolutional neural network (ConvNet or CNN) to individually separate and recombine the style and content of arbitrary images to generate perceptually striking “art” [2]. Given a content and style image as reference, a pre-trained VGG-16 ConvNet can extract feature maps from various layers. Feature maps hold semantic information about both reference images. Loss functions can be developed for content and style by minimizing the mean-square-error between the feature maps used. These loss functions can be additively combined and optimized to render a stylistic image [6]. This technique is called Neural Style Transfer (NST) originally proposed by Leon Gatys in his 2015 research paper, “A Neural Algorithm of Artistic Style”. This research project attempts to replicate and improve upon the work done by Leon Gatys. The purpose of this research is to experiment using a variety of feature maps and optimizing the loss function to identify visually appealing results. A total variation loss factor is introduced to minimize pixilation and sharpen feature formation. Images generated have been assigned a Mean Opinion Score (MOS) by a group of non-bias individuals to affirm the attractiveness of the results.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".