About the role
About the Role
We are building out a computer vision team focused on optimizing on-device face recognition systems, with an emphasis on improving robustness under real-world conditions (occlusion, low-light) and enabling efficient deployment on mobile hardware. This role will work on applied research and engineering across the full pipeline — from model design to on-device deployment.
Responsibilities
- Design and train CNN-based models for face/image recognition, with a focus on robustness to partial occlusion (e.g., masks, glasses, accessories)
- Develop and implement attention-based or occlusion-aware architectures to improve recognition accuracy under obstructed conditions
- Research and build lightweight low-light image enhancement pipelines to improve recognition performance in poor lighting conditions
- Apply model compression techniques (quantization, knowledge distillation) to optimize models for deployment on resource-constrained edge/mobile chipsets
- Benchmark model accuracy, latency, and power consumption across different hardware targets
- Collaborate with cross-functional teams to translate research prototypes into production-ready, on-device solutions
- Stay current with academic and industry developments in face recognition, liveness detection, and efficient deep learning
Requirements
Required Qualifications
- Bachelor's or Master's degree in Computer Science, Electrical Engineering, Machine Learning, or a related field
- Strong hands-on experience with deep learning frameworks (PyTorch or TensorFlow)
- Solid understanding of CNN architectures and computer vision fundamentals
- Experience with at least one of the following:
- Occlusion-robust recognition, attention mechanisms, or face/object recognition under partial visibility
- Low-light image enhancement or image restoration techniques
- Model quantization, knowledge distillation, or other model compression methods
- Experience deploying models on edge devices or mobile chipsets (e.g., mobile SoCs, NPUs) is a strong plus
Preferred Qualifications
- Prior experience with face recognition, biometrics, or liveness/anti-spoofing detection systems
- Familiarity with on-device AI deployment frameworks (e.g., TensorFlow Lite, ONNX Runtime, TensorRT)
- Publications or open-source contributions in computer vision or efficient deep learning
- Experience working with real-time, latency-sensitive applications
Requirements
- Bachelor's or Master's degree in Computer Science, Electrical Engineering, Machine Learning, or a related field