Initial commit

Author: BinaryAlley

.gitignore

@@ -0,0 +1,27 @@
+**/.vs/
+**/.vscode/
+**/.idea/
+**/obj/
+**/out/
+**/build/
+**/.git/
+**/.editorconfig
+**/.fake
+**/__pycache__/
+.externalNativeBuild
+.cxx
+local.properties
+**/.gradle/
+**/gradle/
+src/trainer/ClassifAI_venv/
+src/client/local.properties
+src/client/gradlew
+src/client/gradlew.bat
+src/client/gradle.properties
+src/client/*.iml
+src/client/*.log
+src/client/*.tmp
+src/client/*.bak
+src/client/*.swp
+src/client/*.jks
+src/client/*.keystore

README.md

@@ -0,0 +1,79 @@
+# ClassifAI
+This repository contains a machine learning project focused on recognizing flower species. It consists of two main components:
+
+- A trainer Python application, used for training a TensorFlow Lite Convolutional Neural Network on a dataset consisting of images of five different flower species.
+
+- A client Android application that utilizes the model trained to recognize flower species. It can identify flowers either from static images or live camera feeds.
+
+![ClassifAI](./documents/screenshots/UI_002.png)
+
+## Dependencies
+To run this project, you need:
+* **Python 3+** (optional, for training the neural model - the trained model is already included in this repo, if you want to use it as is and not re-train it from scratch)
+* **TensorFlow Lite library** - For inference on mobile devices using TFLite models built for Android apps that can be developed with Java or Kotlin programming languages, respectively.
+* **Android Studio IDE** - For developing, testing and compiling the Android app.
+
+## Python Model Training
+<p style="text-align: center; background: #000000;">
+  <img src="./documents/screenshots/model_quant.tflite.svg"/>
+</p>
+
+
+The Python script `deep_neural_network_classifier.py` handles training a Deep Neural Network classifier using the [TensorFlow](https://www.tensorflow.org/api_docs/python/tf/) and [Keras APIs](https://keras.io/api/).
+
+
+It downloads and extracts the Flower Photos dataset from TensorFlow, consisting of 3670 photos of 5 flower species - daisies, dandelions, roses, sunflowers and tulips.
+
+
+The images are split 80/10/10 into training, validation and test sets respectively.
+Several data augmentation techniques are applied to the training set, including random rotations, flips and zooms. This expands the effective size and diversity of the training dataset.
+
+
+The model architecture is a Convolutional Neural Network (CNN) with the following layers:
+
+- 3 [Convolution2D](https://keras.io/api/layers/convolution_layers/convolution2d) layers for feature extraction
+- 3 [MaxPooling2D](https://keras.io/api/layers/pooling_layers/max_pooling2d) layers for spatial dimensionality reduction
+- 1 [Dropout layer](https://keras.io/api/layers/regularization_layers/dropout) to prevent overfitting
+- 1 [Flatten layer](https://keras.io/api/layers/reshaping_layers/flatten)
+- 1 [Dense hidden layer](https://keras.io/api/layers/core_layers/dense) with 128 units
+- 1 Output layer with 5 nodes (one per flower class)
+
+<p style="text-align: center;">
+  <img src="./documents/screenshots/model_quant_begin.png"/>
+</p>
+
+
+The model is trained for 15 epochs with a batch size of 32, using the [Adam optimizer](https://arxiv.org/abs/1412.6980) and [Sparse Categorical Crossentropy](https://www.tensorflow.org/api_docs/python/tf/keras/losses/SparseCategoricalCrossentropy) loss function.
+
+Model training and validation plots are generated to visualize accuracy and loss metrics.
+
+The trained model is exported to [TensorFlow Lite format](https://www.tensorflow.org/lite/api_docs), applying [full integer quantization](https://www.tensorflow.org/lite/performance/post_training_integer_quant) for optimized size, suitable for mobile processing power. Metadata including category labels is also embedded.
+
+## Android Recognition App
+
+
+<p style="text-align: center;">
+  <img src="./documents/screenshots/UI_002.png" style="margin: 5px; width: 200px;"/>
+  <img src="./documents/screenshots/UI_003.png" style="margin: 5px; width: 200px;"/>
+  <img src="./documents/screenshots/UI_004.png" style="margin: 5px; width: 200px;"/>
+</p>
+
+
+The Android app provides real-time flower recognition functionality using the TFLite model trained with the Python couterpart application.
+
+The main `RealtimeRecogActivity` activity shows a live camera preview and runs analysis on each frame.
+
+The `ImageAnalyzer` class handles frame processing, invoking the `Classifier` to run inference using the TFLite Interpreter API.
+
+Recognition results with confidence scores are displayed on screen. Photos of recognized flowers can be captured and saved to the device storage.
+
+The app demonstrates how an offline on-device ML model can enable real-time intelligent features, specifically optimized for the computational constraints typical of mobile devices.
+
+<p style="text-align: center;">
+  <img src="./documents/screenshots/Activity Diagram.png"/>
+</p>
+
+
+## License
+
+This project is licensed under the GPLv3.0. See the [LICENSE](./documents/LICENSE.md) file for details.