Microsoft drops Florence-2, a unified model to handle a variety of vision tasks

At present, Microsoft’s Azure AI group dropped a brand new imaginative and prescient basis mannequin known as Florence-2 on Hugging Face.

Accessible below a permissive MIT license, the mannequin can deal with quite a lot of imaginative and prescient and vision-language duties utilizing a unified, prompt-based illustration. It is available in two sizes — 232M and 771M parameters — and already excels at duties equivalent to captioning, object detection, visible grounding and segmentation, acting on par or higher than many giant imaginative and prescient fashions on the market.

Whereas the real-world efficiency of the mannequin is but to be examined, the work is predicted to offer enterprises a single, unified strategy to deal with several types of imaginative and prescient functions. This can save investments on separate task-specific imaginative and prescient fashions that fail to past their major operate, with out intensive fine-tuning.

What makes Florence-2 distinctive?

At present, giant language fashions (LLMs) sit on the coronary heart of enterprise operations. A single mannequin can present summaries, write advertising copies and even deal with customer support in lots of circumstances. The extent of adaptability throughout domains and duties has been superb. However, this success has additionally left researchers questioning: Can imaginative and prescient fashions, which have been largely task-specific, do the identical? 

On the core, imaginative and prescient duties are extra advanced than text-based pure language processing (NLP). They demand complete perceptual capacity. Primarily, to realize common illustration of various imaginative and prescient duties, a mannequin should be able to understanding spatial information throughout completely different scales, from broad image-level ideas like object location, to fine-grained pixel particulars, in addition to semantic particulars equivalent to high-level captions to detailed descriptions.

When Microsoft tried fixing this, it discovered two key roadblocks: Shortage of comprehensively annotated visible datasets and the absence of a unified pretraining framework with a singular community structure that built-in the flexibility to know spatial hierarchy and semantic granularity.

To deal with this, the corporate first used specialised fashions to generate a visible dataset known as FLD-5B. It included a complete of 5.4 billion annotations for 126 million photos, protecting particulars from high-level descriptions to particular areas and objects. Then, utilizing this information, it skilled Florence-2, which makes use of a sequence-to-sequence structure (a kind of neural community designed for duties involving sequential information) integrating a picture encoder and a multi-modality encoder-decoder. This permits the mannequin to deal with numerous imaginative and prescient duties, with out requiring task-specific architectural modifications​​.

“All annotations within the dataset, FLD-5B, are uniformly standardized into textual outputs, facilitating a unified multi-task studying strategy with constant optimization with the identical loss operate as the target,” the researchers wrote within the paper detailing the mannequin. “The end result is a flexible imaginative and prescient basis mannequin able to performing quite a lot of duties… all inside a single mannequin ruled by a uniform set of parameters. Job activation is achieved by means of textual prompts, reflecting the strategy utilized by giant language fashions.”

Efficiency higher than bigger fashions

When prompted with photos and textual content inputs, Florence-2 handles quite a lot of duties, together with object detection, captioning, visible grounding and visible query answering. Extra importantly, it delivers this with high quality on par or higher than many bigger fashions. 

As an example, in a zero-shot captioning check on the COCO dataset, each 232M and 771M variations of Florence outperformed Deepmind’s 80B parameter Flamingo visible language mannequin with scores of 133 and 135.6, respectively. They even did higher than Microsoft’s personal visible grounding-specific Kosmos-2 mannequin.

When fine-tuned with public human-annotated information, Florence-2, regardless of its compact measurement, was capable of compete intently with a number of bigger specialist fashions throughout duties like visible query answering.

“The pre-trained Florence-2 spine enhances efficiency on downstream duties, e.g. COCO object detection and occasion segmentation, and ADE20K semantic segmentation, surpassing each supervised and self-supervised fashions,” the researchers famous. “In comparison with pre-trained fashions on ImageNet, ours improves coaching effectivity by 4X and achieves substantial enhancements of 6.9, 5.5, and 5.9 factors on COCO and ADE20K datasets.”

As of now, each pre-trained and fine-tuned variations of Florence-2 232M and 771M can be found on Hugging Face below a permissive MIT license that permits for unrestricted distribution and modification for industrial use or personal use. 

It is going to be fascinating to see how builders will put it to make use of and offload the necessity for separate imaginative and prescient fashions for various duties. Small, task-agnostic fashions cannot solely save builders the necessity to work with completely different fashions but additionally reduce down the compute prices by a major margin.