台北市南港區5年以上大學
工作內容:
1. 將AI ISP 影像處理模型部署至車用控制器平台(如 Qualcomm、NVIDIA、黑芝麻、MTK及各家NPU 等),確保AI模型在各類平台穩定運行。
2. 依據產品需求,評估模型在 GPU、CPU、NPU 等運算單元上的算力需求與資源佔用,進行性能分析與優化。
3. 與 AI ISP 演算法開發人員合作,完成模型移植、壓縮、加速與平台適配。
4. 針對不同平台環境進行測試與調優,確保模型滿足車用場景的即時性、穩定性與可靠性要求。
5. 參與車載系統軟硬體整合,協助解決模型部署過程中的技術問題。
6. 持續關注車用 AI SoC、NPU 與 ISP 平台的最新技術動態,熟悉各家平台與NPU使用方式,並提出相關改善與升級建議。
技能描述:
1. 資訊工程、電機電子工程、光電工程、影像處理等相關專業碩士及以上學歷。
2. 5年以上AI 模型開發或平台部署經驗。
3. 熟悉深度學習模型部署流程,具備 TensorRT、ONNX Runtime、TVM 或類似工具的經驗。
4. 具備 C/C++、Python 程式開發能力,了解嵌入式 Linux 環境。
Responsibilities:
1. Deploy AI ISP models onto automotive computing platforms (e.g., Qualcomm, NVIDIA, Black Sesame, MTK and NPU-based platforms), ensuring stable runtime performance.
2. Evaluate computational requirements and system resource usage (GPU, CPU, NPU, etc.) based on product needs, and perform resource consumption and performance optimization.
3. Collaborate with AI ISP algorithm developers to handle model porting, compression, acceleration, and platform adaptation.
4. Conduct testing and fine-tuning across different automotive platforms to guarantee real-time, stability, and reliability standards.
5. Support software-hardware integration for automotive systems and troubleshoot deployment-related technical issues.
6. Stay updated with the latest automotive AI SoC, NPU, and ISP platform technologies, providing insights for improvements and upgrades.
Skills Qualification and Experience:
1. Master's degree or above in Computer Science, Electrical/Electronic Engineering, Optoelectronics, Image Processing, or related fields.
2. Minimum of 5 years of experience in AI model development or platform deployment.
3. Familiar with deep learning model deployment workflows, with hands-on experience in tools such as TensorRT, ONNX Runtime, TVM, or similar frameworks.
4. 4Proficient in C/C++ and Python programming, with a solid understanding of embedded Linux environments.