Responsibilities
- Research and implementation of novel/advanced deep neural network architectures and learning techniques to solve a variety of computer vision tasks and push the state of the art in performance.
- The topics may include but not limited to: object/key-point detection, semantic segmentation, face attributes/landmark/recognition, video understanding/behavior modeling, and efficient architecture.
- Problem formulation, benchmarking, dataset analysis/collection, designing and finetuning model with required accuracy and speed.
- Build and maintain the infrastructure for training and deploying models, including data pipelines, experiment management platform, visualization tools, etc.
- Integration of model components with the product stack.
Essential Requirements
- MS / PhD in Computer Science, Electric Engineering, and/or Artificial Intelligence, Machine Learning related technical field with 2+ years of deep learning experience on computer vision.
- 3-5 years of software engineering experience in an academic or industrial society.
- Holistic understanding of deep learning concepts, state of the art in computer vision research and the mathematics of machine learning.
- Proficiency in at least one of the popular computational and deep learning frameworks, such as Pytorch, TensorFlow, Keras, etc.
- Proficiency in Python and C/C++.
- Proven track record of high-quality engineering output (side projects, internships, research projects, full-time jobs etc.).
- The ability and desire to work in the dynamic environment of an early-stage company.
Desirable Pluses
- Research experience in Algorithms, Architecture, Artificial Intelligence, Data Mining, Distributed Systems, Machine Learning, Networking, or Systems.
- A solid foundation in computer science, with competencies in data structures, algorithms, and software design.
- Strength on creativity and communication.
技能標籤: computer vision, deep learning, python
轉職原因:希望能深耕 AI 這一塊。
10 年以上的電腦視覺經驗,目前專研 deep learning。
有多年跨領域的實際經驗,如演算法研究、軟硬體結構優化、嵌入式系統優化等。
因為過去工作公司和 Kneron 有一些合作,因此由用人主管和人選先面談,人選相關經驗較廣,目前單位需要 DL 經驗較資深的人選帶領部門進步。
接著由軟體部門主管、工程師加入面試,主要詢問過去相關經歷,沒有白板題,面試後覺得還不錯,喜歡公司發展方向、產品定位。
專注研究電腦視覺、深度學習等相關問題,擁有傑出的學術背景及相關經驗。
和單位主管、跨部門主管進行面試,總共時間約一小時。
前 30 分鐘進行自我介紹、過去專案介紹,後 30 分鐘為公司介紹,主要的問題都圍繞在相關經歷,沒有白板題,要求解釋 YOLO/HRT/mobile net 的架構等原理。
對於耐能的產品蠻喜歡的,以耐能的 NPU 為主,若對面試評分,滿分 10 分,至少有 8 分,兩位面試主管都很 nice,但因為沒有長期共事過,不確定風格會是如何。
主管評價:極好,將安排跟美國的算法工程師面試,並請準備 DL 基本框架概念等,待 HR 通知。
博士畢,豐富 SoC、ASIC 經驗,半導體產業相關經歷;專注機器學習、算法開發等相關技術超過 7 年。
工作流程和過去很類似,蠻喜歡的,由兩位 Team member 面試,需對 DL 模型建構、DL 細節有一定了解。未來二面需跟美國大主管面試,相關技術會問得更深入。過程很順利,深受面試官喜愛,互動良好,等待後續通知。
碩士畢,6年以上相關經歷,AWB 控制經驗,精通 C++、Python、機器學習框架。
第一關面試:兩小時,與美國團隊主管
整體蠻順利,面試關卡不一定,會依照不同面試走向而有所調整流程,領域算部分符合,今天團隊偏系統相關,是有興趣的。
第二關面試:三小時,兩個美國主管+台灣主管
有簡單聊到薪資規劃,並評估職位長期發展動態、職涯規劃,主管表示接下來由 HR 會跟他談。