Elon Musk’s AI venture, xAI, has introduced an upgraded version of its Grok 1.5 model, now dubbed Grok 1.5 Vision, which incorporates computer vision capabilities. This enhancement allows the model to comprehend and respond to queries about images, marking a significant advancement in its functionality. The unveiling of Grok 1.5 Vision follows closely on the heels of OpenAI’s introduction of GPT-4, which also integrates computer vision features.
The announcement of this development was made through xAI’s official X account, where they detailed the model’s new attributes in a blog post. While the core functionalities of Grok 1.5 remain intact in this latest iteration, the inclusion of vision capabilities is poised to broaden its utility in real-world interactions.
xAI conducted comprehensive benchmark assessments to gauge Grok 1.5 Vision’s performance across various metrics, including their proprietary RealWorldQA benchmark, which evaluates the model’s grasp of real-world spatial concepts. Additionally, the model underwent evaluations in other tests such as MMMU and ChartQA. Notably, Grok surpassed OpenAI’s GPT-4 with Vision and Google’s Gemini 1.5 Pro in the RealWorldQA test, although its performance was comparatively lower in other assessments.
Computer vision stands at the forefront of computer science, aiming to empower computers, including AI models, with the ability to recognize and comprehend real-world objects depicted in images and videos. The objective is to endow machines with vision capabilities akin to those of humans.
Leading technology firms are heavily investing in the development of AI models with vision capabilities. Google’s Gemini 1.5 Pro and OpenAI’s GPT-4 with Vision are notable contenders in this domain.
The potential applications of computer vision are vast and transformative. For instance, Healthify, an Indian platform specializing in calorie tracking and nutrition, recently introduced a feature called ‘Snap’. This feature allows users to capture photos of food items, with the AI then suggesting healthier recipe modifications and exercise plans to balance calorie consumption. Computer vision also holds promise in fields such as medical diagnosis, autonomous vehicles, and beyond.