🔒
Protected Research Portal
Enter access key to view thesis benchmark results.
Unlock Dashboard
Invalid password! Access denied.
Research Portal
Multi-Crop Disease Identification & Visual Benchmarking
Supervisor:
Md. Muksit Ul Islam (Assistant Professor)
Authors:
Md. Rakaatul Islam Ratul, Mehedi Al Asif, Md. Abdullah, Md. Shamim Ahmed
1. Data Preparation
📌 Import, Masking & YOLOv11 Pipeline
(Coming Soon)
2. Benchmarked Models (12)
🧬
Baseline & Modern CNNs
▾
01
CBAM-BasicCNN
8c
Weights
02
ResNet50V2
8c
Weights
03
MobileNetV3-Large
8c
Weights
04
EfficientNetV2-M
9c
Weights
05
ConvNeXt-Tiny
8c
Weights
🔀
Hybrid Architecture
▾
06
CoAtNet-0
8c
Weights
⚡
Vision Transformers (ViT)
▾
07
ViT-B-16
8c
Weights
08
Swin-Tiny
8c
Weights
09
DeiT-III
8c
Weights
🚀
Foundational & Emerging
▾
10
DINOv2
8c
Weights
11
EVA-02-Tiny
8c
Weights
12
SigLIP2
8c
Weights
3. Experimental Trials
🧪 Failed Iterations & Ablation Milestones
(Coming Soon)
Target Cell:
Training Parameters
Model Training
Training Time
Evaluation and Results
Top ↑