MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621081507 A) filed by Symbiosis International Deemed University on July 01, 2026, for An Edge-Enhanced Dual-Branch Cnn With Adaptive Attention For Plant Leaf Disease Detection.
Inventors include Dr. Aniket K. Shahade; and Dr. Priyanka V. Deshmukh.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT AN EDGE-ENHANCED DUAL-BRANCH CNN WITH ADAPTIVE ATTENTION FOR PLANT LEAF DISEASE DETECTION The present invention discloses an edge-enhanced dual-branch CNN framework (100) for plant leaf disease detection. The framework (100) comprises a parallel edge-enhancement branch (120) with learnable Sobel filters (122) and multi-scale edge fusion (124), a spatial attention branch (130) with dynamic channel-spatial attention (132) and background suppression (134), a cross-branch fusion module (140), and a classification head (150). The learnable Sobel kernels adapt to disease-specific edges through backpropagation. The multi-scale fusion captures features across three hierarchical scales. The framework achieves 96.7 percent accuracy with only 3.2 million parameters, enabling edge device deployment for precision agriculture. [
Disclaimer: Curated by HT Syndication.
