Update README.md

README.md

@@ -1,2 +1,7 @@
 # deepstream-yolo-e2e
 Implementation of End-to-End YOLO Models for DeepStream
+
+
+ This repository provides an implementation of End-to-End YOLO models for DeepStream, aiming to streamline the inference process by offloading Non-Maximum Suppression computations onto the YOLO model itself. This design allows users to leverage dynamic batch sizes and input sizes seamlessly.
+
+By eliminating the need for local NMS processing, users gain the flexibility to configure batch sizes and input sizes on-demand without compromising performance. All NMS computations remain integral to the YOLO model, optimizing resource utilization and simplifying model deployment.