【复旦大学研发“眸思”大模型，为视障者开辟“看见”新途径】
上海——复旦大学自然语言处理实验室（FudanNLP）的科研团队近日推出了一项创新成果，名为“眸思”（MouSi）的多模态大模型，旨在帮助视障人士更好地“看见”世界。这一突破性的技术已应用于“听见世界”App，通过一枚普通的摄像头和一对耳机，就能将视觉画面实时转化为语音描述，为视障者提供了一种全新的感知环境的方式。
据复旦大学官方公众号报道，该系统不仅能准确地将图像信息转化为语言，帮助视障用户理解周围环境，还具备描绘场景和提示风险的功能。例如，它可以识别并描述人们、物体、文字，甚至交通信号，从而增强视障人士在日常生活中的行动安全和独立性。
“眸思”大模型的研发，是人工智能技术在无障碍领域的一次重要实践，它将科技的温度传递给需要的人群，让视障者能够更深入地参与社会生活。这一创新成果不仅彰显了复旦大学在人工智能研究领域的领先地位，也体现了科技向善、人文关怀的价值理念。
“听见世界”App的上线，标志着我国在无障碍科技领域的又一重大进展，有望为全球视障者的生活带来深远影响。随着技术的进一步完善和推广，我们期待“眸思”能够为更多视障人士打开一扇通向更广阔世界的窗口。
英语如下：
**News Title:** “Fudan University Develops ‘MouSi’ Large Model, Enabling Visually Impaired to ‘See’ the World Through ‘Hear the World’ App”
**Keywords:** MouSi Large Model, Visual Impairment Assistance, Hear the World App
**News Content:**
**[Fudan University Develops ‘MouSi’ Large Model, Paving a New Path for the Visually Impaired to ‘See’]**
Shanghai – Researchers from the Natural Language Processing Laboratory (FudanNLP) at Fudan University have recently unveiled an innovative project called “MouSi,” a multimodal large model designed to help visually impaired individuals better “see” the world. This groundbreaking technology has been integrated into the “Hear the World” App, converting visual scenes into real-time audio descriptions using a standard camera and a pair of headphones, offering a novel way for the visually impaired to perceive their surroundings.
According to the official Fudan University WeChat account, the system accurately transforms image information into language, assisting visually impaired users in understanding their environment. It also possesses scene depiction and risk-alert capabilities. For instance, it can recognize and describe people, objects, text, and even traffic signals, enhancing the safety and independence of visually impaired individuals in their daily lives.
The development of the “MouSi” large model represents a significant practical application of artificial intelligence (AI) in the field of accessibility, conveying the warmth of technology to those in need. This innovative achievement underscores Fudan University’s leading position in AI research and demonstrates the values of technology for good and人文关怀 (humanistic concern).
The launch of the “Hear the World” App signals another major advancement in China’s accessible technology sector, with the potential to significantly impact the lives of visually impaired individuals worldwide. As the technology continues to evolve and expand, we anticipate that “MouSi” will open a gateway to a broader world for many more visually impaired people.
【来源】https://www.ithome.com/0/753/295.htm
Views: 1