Spectricity 在 X-FAB 部署专有技术,为移动设备带 来光谱成像功能

实现在 CMOS 图像传感器上集成小型像素化光谱滤波器

中国北京,2022 年 6 月 29 日——全球公认的卓越的模拟/混合信号晶圆代
工厂 X-FAB Silicon Foundries(“X-FAB”)今日宣布,与专注多光谱成像技术研
发的半导体设计公司 Spectricity 携手,推出独特成像解决方案。来自 Spectricity
的专有光谱成像技术已部署在 X-FAB 工艺平台的量产设备中;此举将首次实现移
动设备光谱图像传感器的低成本制造。

Spectricity 的专利光谱传感器解决方案将小型像素化的光谱滤波器集成至可
见和近红外光谱范围的 CMOS 图像传感器上,可包含 16 个或更多的高质量窄带
通道。通过此次部署,X-FAB 成为唯一能够批量生产如此小尺寸光谱像素 CMOS
的代工厂,使其和 Spectricity 能够更好地满足移动手持设备 OEM 的需求。由此
带来的解决方案凭借紧凑性和高光谱密度提供了高空间和光谱分辨率感知功能,
助力厂商把握智能手机、物联网、医疗保健和可穿戴设备等领域的机遇。

将光谱相机模块整合至移动设备,贴合了目前增加先进传感功能的持续趋
势。光谱成像仪将能够捕捉多个光谱通道的图像数据,超越传统的红、绿、蓝颜
色通道,并将允许测量物体的光谱特征。需要更为精确自动白平衡的应用,如图
像采集、个性化的化妆品和护肤品、远程医疗及智能园艺/农业等,将会由此受
益。

Spectricity 首席执行官 Vincent Mouret 表示:“我们将最初由 imec 开创的
前沿技术,成功部署至 X-FAB 量产设备中,是 Spectricity 在服务全球智能手机市
场征程中的一个关键里程碑,这也使得我们与实现推动在移动设备中大规模采用
光谱传感的目标更接近。

X-FAB 首席技术官 Jörg Doblaski 评论说:“Spectricity 团队的工程专业技能
给我们留下深刻的印象;我们对与一家创新的欧洲企业成功合作实现光谱成像仪
产品而感到十分兴奋。其技术转让已在很短时间内完成,最初的批次也已达到所
需规格。

Spectricity 计划于 2022 年晚些时候开始推出其由 X-FAB 制造的光谱成像解
决方案样片。

缩略语:
CMOS 互补金属氧化物半导体
IoT 物联网
NIR 近红外线
OEM 原始设备制造商

关于 X-FAB:
X-FAB 是领先的模拟/混合信号和 MEMS 晶圆代工集团,生产用于汽车、工业、消
费、医疗和其它应用的硅晶圆。X-FAB 采用尺寸范围从 1.0μm 至 130nm 的模块化
CMOS 和 SOI 工艺,及其特色 SiC 与微机电系统(MEMS)长寿命工艺,为全球客户打
造最高的质量标准、卓越的制造工艺和创新的解决方案。X-FAB 的模拟数字集成电路
(混合信号 IC)、传感器 MEMS 在德国、法国、马来西亚和美国的六家生产基地生产,
并在全球拥有约 4,000 名员工。www.xfab.com

关于 Spectricity:
Spectricity 是一家总部设在比利时的无晶圆厂公司,其利用 CMOS 技术开发和制造
光谱传感解决方案。这些解决方案专为大批量生产而设计,主要面向消费和移动设备。
公司于 2018 年成立,并得到风险投资的支持。其技术建立在 imec 独家授权的专利组合
基础之上;其中,imec 作为世界领先的纳米电子和数字技术研发与创新中心,其所推动
的技术和产品凝聚了超过 15 年的研究与开发成果。更多信息,请访问:
spectricity.com。


Sr. Computer Vision & AI Engineer

Your main responsibilities will be to
– Develop spectral imaging processing algorithms and calibration techniques to output correct spectral image data from our cameras in volumes
– Support the engineering team to optimize the camera products and production test based on the spectral algorithm requirements
– Together with the application team, develop targeted spectral imaging applications extracting spectral features leaning on computer vision and machine learning modeling
– Co-define and execute the company roadmap

Your profile, skills and talents

Ideally you are a creative, self-dependent engineer or scientist, highly-motivated to be part of a dynamic team creating new value, continuously and actively seeking better solutions for our problems. Specifically:

– You have a master’s degree in mathematics, computer science, engineering or similar relevant degree by proven experience and at least 5-10 years of experience in a computer vision and image processing.
– Signal processing and computer vision are your trade. You are comfortable at all levels of abstraction, from architecture to implementation of algorithm code. You are on top of quality standards with respect to software implementation and verification
– You have proven experience in applied Machine Learning and development of AI-based solutions.
– You are proficient with Python. Git and continuous integration and unit tests are the tools you use daily
– Experience with spectroscopic analysis is a very strong plus
– You are a leader who can efficiently interact with your peers within CTO office, engineering and marketing teams
– You are fluent in English, and have good presentation skills

You will help us be successful when:
– You are a good team player, able to pursue and find compromise through constructive discussion, enthusiasm and skills
– You are result-driven, you have a natural desire to analyze and solve problems and you are self-critical


spectricity Image Sensors Europe

Senior System Engineer

Your main responsibilities will be 

  • Understand the end product use case in the customer’s system
    • Understand the target system components and interfaces
    •  Understand the customer needs to easy design in of our products
    •  Understand the eco-system of the application
  •  Translate this information into product requirements to meet the customer needs
  •  Architect and simulate the optical-electrical system in its various components (lens, filters, sensor).
  •  Implement the chosen solution:
    • Design (Define DOE when needed) & tape-out
    • Results analysis and debugging
    • Fine tune for manufacturing
  • Interface to internal and external customers to align needs and development specifications
  • Give inputs to the technical roadmap to define future products
  • Hands-on system modelling and simulations.

In your role, you

  • Are part of the Engineering team, focusing on product architecture and development execution of the product in its different components, reporting to VP Engineering.
  • Will collaborate closely with the CTO office, process technology, R&D, Business Development and operations teams.
  • Collaborate closely with our customers and suppliers.

Your profile, skills and talents

  • You have a master’s degree in engineering, sciences or a similar technical discipline; a PhD degree is a plus.
  • At least 10 years of hands-on industry experience with:
    • Image sensor design and applications
    • Optical systems (lens, filters)
    • Good understanding of the semiconductor development flow
    • Image processing data path knowledge is a plus
  •  You are fluent in English and have excellent communication skills.

You will help us be successful when:

  • You are an excellent team player, able to pursue and find compromise through constructive discussions,
    enthusiasm and skills.
  • You are result-driven, you have a natural desire to analyze and solve problems and you are self-critical.
  • You are highly focused on delivering and finding solutions in a complex environment with many stakeholders and trade-offs.
  • You have a high sense of urgency to implement quality principles and procedures.

Senior AI expert

Your main responsibilities will be 

  • Develop AI algorithms in our application and computer vision team, for our spectral imaging systems for consumer-oriented applications;
  • Steer the engineering team to optimize the camera products and production test based on the application requirements;
  • Co-define and execute the company AI and application roadmap.

Your profile, skills and talents

Ideally you are creative and self-dependent, highly-motivated to be part of a dynamic team creating new value, continuously and actively seeking better solutions for our problems. Specifically:

  • You have a master’s degree in mathematics, computer science, engineering or similar relevant degree by proven experience and at least 5-10 years of experience in AI.
  • AI is your trade. You have many years of proven industrial experience in proposing and developing neural network architectures for real world applications. You have a deep understanding of the architectures and their merits and issues, beyond simple use of neural nets.
  • You are able to propose and develop training routines and merit functions, understand requirements on generating the appropriate large training datasets, and you have a critical mindset to analyze results and iterate forward.
  • You are proficient with Python, Keras, Tensorflow, and preferably also with Pytorch, VertexAI and GCP. Git and continuous integration and unit tests are the tools you use daily.
  • Experience with embedded implementations on hardware and edge compute is a strong plus.
  • Experience with computer vision and/or spectroscopy is a plus.
  • You are a leader who can efficiently interact with your peers within CTO office, engineering and marketing teams.
  • You are fluent in English, and have good presentation skills.

You will help us be successful when:

  • You are a good team player, able to pursue and find compromise through constructive discussion, enthusiasm and skills.
  • You are result-driven, you have a natural desire to analyze and solve problems and you are self-critical.