computer vision projects

watermark detection siglip2 is a vision language encoder model fine-tuned from siglip2 patch16 224. it detects whether an image contains a watermark or not, using the siglip architecture.

flux midjourney mix2 lora is a text-to-image lora model designed for high-quality visual generation, particularly excelling in realism, close-up shots, and stylized photography, with a base model of flux.1-dev

facial emotion detection siglip2 is an image classification vision-language encoder model fine-tuned from siglip2 patch16 224. it is designed to classify facial emotions

deepfake detector model v1 is a vision-language encoder model fine-tuned from siglip2 512 for deepfake image classification, to detect whether an image is real or generated using synthetic media techniques.

age classification siglip2 is an image classification vision-language encoder model fine-tuned from siglip2 patch16 224. it is designed to predict the age group of a person.

gender classifier mini is an image classification vision-language encoder model fine-tuned from google/siglip2-base-patch16-224. it is designed to classify images based on gender

canopus lora flux ultrarealism-2.0 is a lora-based text-to-image model built on the flux.1-dev base, designed to generate ultra-realistic images with enhanced facial detail and photorealistic quality.

flux super realism lora is a lora-based text-to-image model built on the flux.1-dev base, designed to generate ultra-realistic images with enhanced facial detail and photorealistic quality.

face mask detection is a binary image classification model based on siglip2 patch16 224, trained to detect whether a person is wearing a face mask or not.

realistic gender classification is a binary image classification model based on siglip2 patch16 224, designed to classify gender from realistic human portrait images.

(multilingual vision-language encoders) the model is trained to identify and categorize content types in images, especially for explicit, suggestive, or safe media filtering.

dog breed 120 is an image classification vision-language encoder model fine-tuned from siglip2 patch16 224. it is designed to classify dog images into specific breed categories using the siglip architecture.

bird species classifier 526 is an image classification vision-language encoder model fine-tuned from siglip2 patch16 224. it is designed to classify different bird species using the siglip.