Upload folder using huggingface_hub

README.md

@@ -5,9 +5,24 @@ tags:
 - image-classification
 ---
 
-# Kindwise Crossroad Model
+# Kindwise Router Classifier
 
-## How to use
+This model classifies images based on their content, acting as a "router" to
+direct requests to the correct Kindwise service.  It automatically detects
+whether an image contains human, insect, mushroom, or plant.
+
+Th model is intended to be the first step in an image processing pipeline.
+Instead of having each specialized service (e.g., insect, plant classification)
+analyze every image, this model quickly determines the image's category. This
+reduces latency and optimizes system resources.
+
+## Technical Details and Formats
+
+The modek is available in two optimized formats for easy deployment:
+- **TorchScript**: Optimized for production environments and server-side applications where performance and low latency are critical.
+- **TensorFlow Lite**: Perfect for mobile devices and edge computing, where efficiency and minimal model size are key.
+
+## Usage
 
 Here is how to use this model to classify an image into one of the basic classes:
 
@@ -22,8 +37,8 @@ import torch
 import torchvision
 
 DEVICE_NAME = 'cuda:0'
-MODEL_PATH = hf_hub_download('kindwise/crossroad.tiny', 'model.traced.pt')
-CLASSES_PATH = hf_hub_download('kindwise/crossroad.tiny', 'classes.txt')
+MODEL_PATH = hf_hub_download('kindwise/router.tiny', 'model.traced.pt')
+CLASSES_PATH = hf_hub_download('kindwise/router.tiny', 'classes.txt')
 IMAGE_PATH = '/tmp/photo.jpg' 
 
 with open(CLASSES_PATH) as f:
@@ -76,8 +91,8 @@ from huggingface_hub import hf_hub_download
 import numpy as np
 import tensorflow as tf
 
-MODEL_PATH = hf_hub_download('kindwise/crossroad.tiny', 'model.tflite')  # or model.optimized.tflite
-CLASSES_PATH = hf_hub_download('kindwise/crossroad.tiny', 'classes.txt')
+MODEL_PATH = hf_hub_download('kindwise/router.tiny', 'model.tflite')  # or model.optimized.tflite
+CLASSES_PATH = hf_hub_download('kindwise/router.tiny', 'classes.txt')
 
 with open(CLASSES_PATH) as f:
     CLASSES = [line.strip() for line in f]