2 Gradcam Jobs

As the AI Model Development & Deployment Lead, your primary responsibility will be to design and implement cutting-edge computer vision models for various tasks such as object detection, tracking, segmentation, and action recognition. You should have expertise in architectures like YOLO (v4, v5, v8), Vision Transformers, Mask R-CNN, Faster R-CNN, LSTMs, and Spatio-Temporal Models for in-depth image and video analysis. It will be crucial to contextualize the models for challenging scenarios like poor detection due to occlusion or small, fast-moving objects by training them on diverse situations. Your role will involve identifying reasons for suboptimal model performance and developing a scalable computer vision layer adaptable to different environments. Collaborating with data scientists using frameworks like TensorFlow and Pytorch, you will drive impact across key KPIs and optimize models for real-time inference through techniques such as quantization, pruning, and model distillation. In the realm of Reinforcement Learning & Model Explainability, you will be tasked with developing and integrating reinforcement learning models to enhance decision-making in dynamic AI environments. This will involve working with Deep Q-Networks (DQN), Proximal Policy Optimization (PPO), A3C, SAC, and other RL techniques. Furthermore, you will explore explainability methods such as SHAP, LIME, and Grad-CAM to ensure transparency and interpretability of AI-driven decisions. Additionally, engineering context-aware features from video sequences to extract temporal and spatial insights for improved model explainability and decision accuracy will be part of your responsibilities. Leading the charge in creating a Scalable Data Model for AI Model Training Pipelines, you will oversee the end-to-end data preparation process, including data gathering, annotation, and quality review before training data utilization. The focus will be on enhancing the quality and volume of training data continuously to boost model performance iteratively. Your expertise in planning the data model for future scalability, with the ability to incorporate new data elements in a modular manner, will be essential. Key Technical Skills required for this role include a proven track record in delivering AI products, hands-on experience with advanced AI models, proficiency in managing and mentoring a small team of data scientists, expertise in AI/ML frameworks like TensorFlow, PyTorch, and Keras, effective collaboration with cross-functional teams, and strong problem-solving skills in image and video processing challenges. Nice to have skills include a mindset for experimentation, iterative improvement, and testing to enhance model performance continually, staying updated with the latest AI research, and balancing experimentation with timely product delivery. Collaboration with engineering and product teams to align AI solutions with product goals and deployment of AI models into production are crucial aspects. Additionally, leadership in guiding and mentoring a team of AI scientists, taking ownership of product KPIs related to AI models, and fostering an AI-first culture focused on innovation and continuous improvement will be part of your role.,

Visionify is dedicated to leveraging the potential of Computer Vision and AI for various real-world applications. We are currently seeking a highly skilled, motivated, and enthusiastic Senior Computer Vision Engineer to play a crucial role in implementing our strategic plans. As a Senior Computer Vision Engineer at Visionify, you will be tasked with tackling cutting-edge challenges in the realm of Computer Vision by devising innovative algorithms and optimizations. The majority of our projects revolve around practical applications of Computer Vision, necessitating a strong grasp of contemporary model types such as Classification, Object detection, Object Recognition, OCR, LayoutML, and GAN networks. Proficiency in Pytorch is essential for this role, as it serves as our primary programming language. Familiarity with Azure and Azure ML Studio would be advantageous. Candidates applying for this position should remain abreast of the latest advancements and actively contribute to enhancing the Pytorch project's performance and accuracy. Your expertise in PyTorch and its underlying mechanisms will be pivotal in resolving customer challenges and offering valuable insights into product improvements. Experience in optimizing and streamlining models for deployment on edge devices, as well as converting models to NVIDIA TensorRT, will be highly valued. A strong foundation in Python programming is indispensable, given its widespread use in our organization for developing training and inference pipelines. Effective communication and presentation skills are also crucial. The ideal candidate will exhibit a deep passion for artificial intelligence and a commitment to staying updated on industry trends. **Responsibilities:** - Understanding business objectives and devising Computer Vision solutions that align with these goals, including developing training and inference frameworks and leveraging various ML technologies. - Building and optimizing Pytorch models for different runtime environments, including NVIDIA Jetson TensorRT. - Guiding the development team, addressing their queries, and facilitating the timely completion of their tasks. - Creating ML/Computer Vision algorithms to address specific challenges. - Analyzing and visualizing data to identify potential performance-affecting disparities in data distribution, especially when deploying models in real-world scenarios. - Establishing processes for core team operations, such as data acquisition, model training, and prototype development. - Identifying and utilizing open-source datasets for prototype building. - Developing pipelines for data processing, augmentation, training, inference, and active retraining. - Training models, fine-tuning hyperparameters, and devising strategies to address model errors. - Deploying models for production use. **Requirements:** - Bachelor's or Master's degree in Computer Science, Computer Engineering, IT, or a related field. - Minimum of 5 years of relevant experience; candidates with exceptional skills but less experience are encouraged to apply. - Industry experience in Image & Video Processing, including familiarity with OpenCV, GStreamer, TensorFlow, PyTorch, TensorRT, and various model training/inference techniques. - Proficiency in deep learning classification models (e.g., ResNet, Inception, VGG) and object detection models (e.g., MobileNetSSD, Yolo, FastRCNN, MaskRCNN). - Strong command of Pytorch, Torchvision, and the ability to develop training routines and update models effectively. - Familiarity with Colab, Jupyter Notebook, CUDA/GPU, and CNN visualization techniques like CAM and GradCAM. - Expertise in Computer Vision and real-time video processing methods. - Proficient in Python programming and adept at writing reusable code. - Experience with OpenCV, Scikit packages, NVIDIA platform tools (e.g., Deepstream, TensorRT), Python web frameworks (e.g., Flask, Django, FastAPI), and ML platforms (e.g., PyTorch, TensorFlow). - Knowledge of AWS SageMaker, various databases (e.g., Elasticsearch, SQL, NoSQL, Hive), cloud environments (preferably AWS) for software development, GPU-based training infrastructures, Docker, and DevOps and MLOps best practices for ML systems. **Desired Traits:** - Collaborative mindset and ability to thrive in a team environment. - Adaptability to evolving requirements. - Proclivity for innovative problem-solving. - Strong focus on work quality and developing robust code.,