|
import mxnet as mx |
|
import matplotlib.pyplot as plt |
|
import numpy as np |
|
from collections import namedtuple |
|
from mxnet.gluon.data.vision import transforms |
|
from mxnet.contrib.onnx.onnx2mx.import_model import import_model |
|
import os |
|
import gradio as gr |
|
|
|
mx.test_utils.download('https://s3.amazonaws.com/model-server/inputs/kitten.jpg') |
|
|
|
mx.test_utils.download('https://s3.amazonaws.com/onnx-model-zoo/synset.txt') |
|
with open('synset.txt', 'r') as f: |
|
labels = [l.rstrip() for l in f] |
|
|
|
os.system("wget https://github.com/onnx/models/raw/main/vision/classification/alexnet/model/bvlcalexnet-12.onnx") |
|
|
|
|
|
|
|
sym, arg_params, aux_params = import_model('bvlcalexnet-12.onnx') |
|
|
|
Batch = namedtuple('Batch', ['data']) |
|
def get_image(path, show=False): |
|
img = mx.image.imread(path) |
|
if img is None: |
|
return None |
|
if show: |
|
plt.imshow(img.asnumpy()) |
|
plt.axis('off') |
|
return img |
|
|
|
def preprocess(img): |
|
transform_fn = transforms.Compose([ |
|
transforms.Resize(256), |
|
transforms.CenterCrop(224), |
|
transforms.ToTensor(), |
|
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]) |
|
]) |
|
img = transform_fn(img) |
|
img = img.expand_dims(axis=0) |
|
return img |
|
|
|
def predict(path): |
|
img = get_image(path, show=True) |
|
img = preprocess(img) |
|
mod.forward(Batch([img])) |
|
|
|
scores = mx.ndarray.softmax(mod.get_outputs()[0]).asnumpy() |
|
|
|
scores = np.squeeze(scores) |
|
a = np.argsort(scores)[::-1] |
|
results = {} |
|
for i in a[0:5]: |
|
results[labels[i]] = float(scores[i]) |
|
return results |
|
|
|
|
|
if len(mx.test_utils.list_gpus())==0: |
|
ctx = mx.cpu() |
|
else: |
|
ctx = mx.gpu(0) |
|
|
|
mod = mx.mod.Module(symbol=sym, context=ctx, label_names=None) |
|
mod.bind(for_training=False, data_shapes=[('data', (1,3,224,224))], |
|
label_shapes=mod._label_shapes) |
|
mod.set_params(arg_params, aux_params, allow_missing=True, allow_extra=True) |
|
|
|
title="SqueezeNet" |
|
description="SqueezeNet is a small CNN which achieves AlexNet level accuracy on ImageNet with 50x fewer parameters. SqueezeNet requires less communication across servers during distributed training, less bandwidth to export a new model from the cloud to an autonomous car and more feasible to deploy on FPGAs and other hardware with limited memory." |
|
|
|
examples=[['catonnx.jpg']] |
|
gr.Interface(predict,gr.inputs.Image(type='filepath'),"label",title=title,description=description,examples=examples).launch(enable_queue=True) |