Intellectual Property India Publishes Patent Application for 'A Hybrid Deep Learning Framework For Real-Time Temporal Object Detection Using Yolov4, Vgg16, And Ssd-Rnn Integration' Filed by Mr. Nilesh Kumar Sen; Mr. Dhruv Bhadauria; Mr. Aarav Rathore; and Mr. Aashish Tomar

MUMBAI, India, June 16 -- Intellectual Property India has published a patent application (202611053943 A) filed by Mr. Nilesh Kumar Sen; Mr. Dhruv Bhadauria; Mr. Aarav Rathore; and Mr. Aashish Tomar, Ghaziabad, Uttar Pradesh, on April 28, for 'a hybrid deep learning framework for real-time temporal object detection using yolov4, vgg16, and ssd-rnn integration.'

Inventor(s) include Mr. Nilesh Kumar Sen; Mr. Dhruv Bhadauria; Mr. Aarav Rathore; and Mr. Aashish Tomar.

The application for the patent was published on June 5, under issue no. 23/2026.

According to the abstract released by the Intellectual Property India: "This disclosure presents a hybrid deep learning architecture designed for high-speed, high-accuracy real-time object detection in dynamic video streams. The system integrates a YOLOv4 framework for rapid localization, a VGG16 backbone for robust feature extraction, and an SSD-RNN mechanism to maintain temporal consistency across sequential frames. Trained on the COCO dataset using TensorFlow, the model achieves a optimized trade-off between inference latency and detection precision. Experimental results demonstrate superior performance in urban environments, effectively identifying multi-scale objects such as pedestrians and vehicles at enhanced frames-per-second (FPS) rates. KEYWORDS Real-time Object Detection, YOLOv4, VGG16, SSD-RNN, Temporal Feature Aggregation, Deep Learning, Computer Vision, Autonomous Surveillance."

Disclaimer: Curated by HT Syndication.