While you’ll likely see this roadmap change based on feedback, in the meantime we hope this shows you the direction we’re taking the product. We aim to release most of these improvements within the next few months, although there’s not a firm timeline. We’re also improving performance, reducing memory, and fixing bugs every release. Improve subject detection model accuracy to reduce how often you need to refine inaccurate subjects.Add cropping and allow optional auto-upscaling to original image dimensions.New Sharpen model for focus and motion blur that creates fewer artifacts.New Remove Noise model for non-raw images that preserves more detail and responds better to slider changes.Add Configure Autopilot option to enable or disable Remove Noise at various noise levels, similar to the current Sharpen option.For raw images: improve strong raw noise parameter estimation and allow Autopilot to select it.Further improve subject refinement based on AI Brush feedback.Improve blur estimation so Autopilot can enable or disable Sharpen more reliably.In internal testing, this new model significantly outperforms our existing upscaling models in both Gigapixel and Topaz Photo AI. ![]()
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |