Android Inferener APK - Inferncing a supplied TF Lite Model with Supplied Images on Android Device

I am looking for an experienced Android developer who can create an APK for inferencing a supplied TF Lite model with supplied images on a tablet. The ideal candidate should have experience in Java programming language and have a ready-to-use TF Lite model for inferencing. The project requires the following skills and experience: - Experience in developing APKs for Android devices - Strong knowledge of Java programming language - Familiarity with TF Lite models and inferencing - Ability to work with supplied images and integrate them into the APK The project deliverables should include a fully functional APK for inferencing the supplied TF Lite model with the supplied images on a tablet. The code should be well-documented and easily maintainable. The project timeline is flexible but should be completed within a reasonable timeframe. Android TF Lite Inferncer APK Android TensorFlow Lite Image Classification Inferencing APK Written in Java for 8-bit Images We will supply required image classification model and test images for inferencing Requirements: a) Produce working source code in Java that can be compiled in Android Studio and deployed to an Android device as an APK. b) The code will use TensorFlow Lite to access a folder that has some test images and perform inference on all of them and display the prediction values on the screen. Supplied: 1. A trained quantized TFLite model, “[login to view URL]” 2. 8-bit single-channel test images in the “image” folder 3. Some Java sample code, “[login to view URL]” that does not work. There is an issue loading the images. 4. Some Python code, “[login to view URL]”, that does work 5. The directory where the test images will be read from, and any output can be placed in: “/DCIM/inference” 6. A link to access a folder with all of these items: link Additional Information: The provided images are single-channel 8-bit images. The TFLite model input is a 32f array of pixels so the images will need to be converted before passing them to the model. Additionally, the model was trained with 0 – 255 integer values so the values will need to be converted to float values by multiplying the pixel values by 255.0f before passing them to the model for inference.