TensorFlow Image Detection, Classifier & TFLite
Budget: ₹12,500 – ₹37,500 INR
I already have a folder of labeled images and need a fast turnaround on an image classification pipeline. Your job is to build a high-accuracy model in TensorFlow, validate it, then export a fully functional TensorFlow Lite file suitable for mobile or embedded deployment.
Here’s what I expect: you’ll handle any necessary preprocessing, choose or fine-tune an architecture that balances accuracy with size (e.g., EfficientNet-Lite, MobileNet, or whatever you feel is best), train until results are reliable, and provide clear metrics so I can see the improvement at each stage. Once the model meets the agreed accuracy, convert it to .tflite with all optimizations enabled (quantization preferred), test that the output signatures work, and hand over the .tflite file, the label map, and the training/inference scripts so I can reproduce or update the model later.
Timing is critical—I need the finished .tflite model ASAP—so please outline how quickly you can begin, your expected training schedule, and any questions you have about the dataset as soon as we start.
Here’s what I expect: you’ll handle any necessary preprocessing, choose or fine-tune an architecture that balances accuracy with size (e.g., EfficientNet-Lite, MobileNet, or whatever you feel is best), train until results are reliable, and provide clear metrics so I can see the improvement at each stage. Once the model meets the agreed accuracy, convert it to .tflite with all optimizations enabled (quantization preferred), test that the output signatures work, and hand over the .tflite file, the label map, and the training/inference scripts so I can reproduce or update the model later.
Timing is critical—I need the finished .tflite model ASAP—so please outline how quickly you can begin, your expected training schedule, and any questions you have about the dataset as soon as we start.