Clarifai Release 8.0

The Clarifai platform has been updated to Release 8.0, introducing advanced models for image and video captioning, as well as handwritten optical character recognition (OCR). Key additions include the Multilingual OCR "language-aware-multilingual-ocr-multiplex" model, which employs an end-to-end approach to handle multiple languages and scripts, outperforming existing models on benchmarks like MLT17 and MLT19. The "paddleocr-multilingual" model aims to provide high-quality, multilingual OCR tools, offering lightweight and server series models supporting up to 80 languages. Additionally, the "general-english-image-caption-clip" model leverages OpenAI's CLIP to learn visual concepts via natural language supervision, while Microsoft's TrOCR model enhances handwritten OCR capabilities through a fine-tuned encoder-decoder architecture.

No tracked trend matches for this post yet.