ABDULHAMEED IDRIS ADEDAMOLA

CNN-Based Biometric Recognition Blog CNN-Based Biometric Recognition Using Palmprint and Fingerprint Images Biometric recognition systems are widely used in modern security because they provide reliable identity verification based on unique human traits such as fingerprints and palmprints. However, unimodal systems often suffer from limitations such as noise sensitivity, variability in data, and reduced performance under poor image quality. Approach This work presents a Convolutional Neural Network (CNN) approach using two biometric modalities: palmprint and fingerprint images. The objective is to develop and evaluate independent CNN models for each modality as a foundation for future multimodal biometric fusion. Datasets and Preprocessing The BMPD dataset was used for palmprint recognition, while the FVC2004 DB1-B dataset was used f...

Modality-Adaptive Brain Tumor Segmentation Modality-Adaptive Brain Tumor Segmentation in Medical AI Introduction Brain tumor segmentation is a key task in medical image analysis where AI models are used to detect and outline tumor regions in MRI scans. These systems often rely on multiple MRI modalities such as T1, T2, and FLAIR to achieve high accuracy. Challenge in Real Clinical Environments In real-world healthcare settings, MRI data is often incomplete. Not all imaging modalities are available for every patient. This creates a major challenge because most deep learning models depend on full multi-modal input. Need for Adaptive Models To address this limitation, modern AI systems are being designed to work with partial or single-modality MRI inputs. These models aim to maintain stable performance even when some data sources are missing. Work with single MRI modality input Handle any combination of available modalities Maintain consistent segmentation ...

Feature Selection Using GA, PSO, and ACO Feature Selection Using Metaheuristic Algorithms: GA, PSO, and ACO 1. Introduction In machine learning, datasets often contain many features, but not all of them are useful. Feature selection helps identify the most relevant variables that improve model performance while reducing noise and complexity. 2. What is Feature Selection? Feature selection is the process of selecting a subset of important variables from a dataset. It helps improve accuracy, reduce overfitting, and lower computational cost. 3. Wrapper-Based Feature Selection All three methods discussed (GA, PSO, ACO) are wrapper-based approaches. This means a machine learning model is used to evaluate each feature subset using a performance metric such as AUC-ROC. 4. Genetic Algorithm (GA) Genetic Algorithm is inspired by natural evolution. It uses selection, crossover, and mutation to evolve better feature subsets over generations. 5. Particle Swarm ...

Python Interpreter vs Compiler Explained Python Interpreter vs Compiler Explained Python is often described as an interpreted language, but the way it works is a bit more advanced than a simple interpreter model. 🐍 Python Interpreter Python executes code using an interpreter . This means: Code is executed line by line Errors appear during execution Programs run directly without full pre-compilation When you run a Python file, the interpreter reads and executes it immediately. ⚙️ Compiler (Traditional Meaning) A compiler works differently: It converts the entire program into machine code at once Execution happens after compilation Common in languages like C and C++ 🧠 How Python Really Works Python actually uses a hybrid approach: Your code (.py file) Compiled into bytecode (.pyc) Executed by the Python Virtual Machine (PVM) So, Python combines both compilation and interpretat...

Optimizing Deep Learning Libraries for Edge-AI on Mobile GPUs Optimizing Deep Learning Libraries for Edge-AI on Mobile GPUs ⚡ Edge-AI performance is not just about models — it’s about libraries. Deploying Deep Learning (DL) models on edge devices is constrained by compute, memory, and energy efficiency. On mobile GPUs, performance often depends more on backend optimization than model architecture. Key Libraries cuBLAS (CUDA Basic Linear Algebra Subprograms) cuDNN (CUDA Deep Neural Network library) TensorRT (Tensor Runtime) from NVIDIA Key Insight There is no universal best library. Performance depends on: Input size Model type (CNN vs Vision Transformer) Layer configuration Most deep learning workloads ultimately rely on matrix operations (GEMM), making low-level optimization critical. Takeaway ...

🚀 How I Successfully Submitted My App to Google Play Review (Step-by-Step Process) I just submitted my app update Hierarchy Star Reports v1.4 to Google Play, and it is now officially in review on Google Play Console. Here’s the exact step-by-step process I followed (my deployment methodology / “mergogoly” approach): 🧩 Step 1: Fixed app identity issues Ensured app name matches everywhere (Play Console + strings.xml) Verified no old names like “DALogos” remained in the project 🎨 Step 2: Fixed app icon mismatch Confirmed launcher icon in mipmap/ matches Play Store icon Checked both ic_launcher and ic_launcher_round Ensured installed app icon matches store listing 🔢 Step 3: Updated versioning Increased versionCode (4 → 5) Updated versionName (1.3 → 1.4) Prevented duplicate build upload errors 🔐 Step 4: Proper signing configuration Used correct release keystore (upload-key.jks) Built a signed release AAB 📦 Step 5: Built ...