2 hours ago
Base Salary
$180k - $220k/yr
Responsibilities
- Design and train deep neural networks that perform object detection and tracking simultaneously.
- Evaluate state-of-the-art research papers, prototype their concepts, and adapt them into production-grade solutions.
- Modify model architectures, internal layers, data flows, and custom loss functions through extensive experimentation.
- Develop data augmentation strategies and training and post-training approaches for data-constrained environments.
- Optimize models for accurate, efficient, real-time inference and on-device deployment.
Requirements
- 5+ years of proficiency with Python and PyTorch.
- 3+ years of proficiency with C++ for production deployment and optimization.
- Deep theoretical and practical understanding of object detectors such as Transformers, YOLO variants, and R-CNNs, along with tracking algorithms such as DeepSORT, Kalman Filters, and Optical Flow.
- Demonstrated ability to modify model architectures rather than relying on out-of-the-box APIs.
- Experience improving model generalization with limited data through Transfer Learning, Domain Adaptation, or Few-Shot Learning.
- Strong understanding of linear algebra and probability for custom loss functions and geometric 3D vision.
- Preferred: hands-on experience with 3D point-cloud or LiDAR data.
- Preferred: experience with TensorRT, ONNX Runtime, or edge-specific hardware such as NVIDIA Jetson.
Benefits
- The role may be eligible for equity and benefits.
- Base pay is listed at $180,000-$220,000 per year.
Categories
