Since the rest of the hand’s color saturation was not increased by Gigapixel, the orange stood out significantly. The original photo has a slight orangish hue on the tip of the thumb, which Gigapixel AI detected and increased the saturation, turning the orange brighter. One small visual glitch visible in the processed image was a little discoloration on the thumb. This zoom-in demonstrates how each of the three upscalers handled hair, as the material of the backpack, including the straps. Here are the results for all three upscalers and how they processed hair particles.Īnother zoom-in demonstrates the finer detail of the material, such as the backpack and the fluffy sweater.Īs with hair, fluffy material can be a challenge to upscale due to the smaller particles. Upscaling images with hair is quite hard since hair particles tend to be very tiny. The original image was 360x240 and after being processed through Let’s Enhance, Gigapixel, and waifu2x, the image size was doubled to 720x480. Here are the results of all three upscalers. Even though the camera focus was on the person in the photo, the quality of the background was taken into account as well. ![]() This demonstrates how the software upscale faces, as well as detect other small details, such as hands and hair. ![]() Test #1 was conducted to see how Let’s Enhance, Gigapixel, and waifu2x could enhance a portrait photo. Tests: compare results side-by-sideįour different image types were tested on Let’s Enhance, Gigapixel, and waifu2x. ![]() However, these are vastly different tools and we are going to see how each of them can be used effectively in their own right. Each tool served its purpose and was successfully able to upscale the photos to at least twice their size, without blur or pixelation in the end results. For this comparison, we tested a variety of photos on the most popular AI upscalers on the market: Let’s Enhance, Gigapixel, and waifu2x.
0 Comments
Leave a Reply. |