#!/usr/bin/env python3 # Copyright 2023-2024,2026 NXP # #Licensed under the Apache License, Version 2.0 (the "License"); #you may not use this file except in compliance with the License. #You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # #Unless required by applicable law or agreed to in writing, software #distributed under the License is distributed on an "AS IS" BASIS, #WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. #See the License for the specific language governing permissions and #limitations under the License. import argparse import numpy as np import tensorflow as tf from PIL import Image from labels import IMAGENET_LABELS INPUT_SIZE = 224 # The EfficientNet-lite TFLite model outputs 1001 classes: index 0 is a # "background" class, so the 1000 ImageNet labels are offset by 1. LABEL_OFFSET = 1 def inference(model_path, input_path, output_path=None, top_k=5): interpreter = tf.lite.Interpreter(model_path=model_path) interpreter.allocate_tensors() input_details = interpreter.get_input_details() output_details = interpreter.get_output_details() print("Loaded model:", model_path) img = Image.open(input_path).convert("RGB") img = img.resize((INPUT_SIZE, INPUT_SIZE)) img = np.asarray(img) input_dtype = input_details[0]["dtype"] if input_dtype == np.uint8: # uint8-input model: feed raw [0, 255] pixels. tensor = img.astype(np.uint8) elif input_dtype == np.int8: # int8-input model: shift the [0, 255] pixel range to [-128, 127]. tensor = (img.astype(np.int32) - 128).astype(np.int8) else: # float32-input model: EfficientNet-lite expects pixels in [-1, 1]. tensor = (img.astype(np.float32) / 127.5) - 1.0 tensor = np.expand_dims(tensor, axis=0) interpreter.set_tensor(input_details[0]["index"], tensor) interpreter.invoke() out = interpreter.get_tensor(output_details[0]["index"]) # Dequantize the output if the model returns a quantized tensor. out_scale, out_zero_point = output_details[0]["quantization"] if out_scale != 0: out = out_scale * (out.astype(np.float32) - out_zero_point) probs = out[0].astype(np.float32) top = probs.argsort()[::-1][:top_k] print("Top {} predictions:".format(top_k)) lines = [] for rank, idx in enumerate(top): label = IMAGENET_LABELS.get(int(idx) - LABEL_OFFSET, "unknown") line = " {}. [{}] {} (score: {:.4f})".format(rank + 1, idx, label, probs[idx]) print(line) lines.append(line) if output_path: with open(output_path, "w") as f: f.write("\n".join(lines) + "\n") print("Wrote predictions to", output_path) if __name__ == "__main__": parser = argparse.ArgumentParser(description="EfficientNet-lite0 inference example") parser.add_argument( "-m", "--model", default="assets/quantized_tflite_model/efficientnet-lite0-int8.tflite", help="Path to the TFLite model file.", ) parser.add_argument( "-i", "--input", default="example_input.jpg", help="Path to the input image.", ) parser.add_argument( "-o", "--output", default=None, help="Optional path to write the top predictions as text.", ) args = parser.parse_args() inference(args.model, args.input, args.output)