Spaces:
Sleeping
Sleeping
| import fastai | |
| import PIL | |
| import gradio as gr | |
| import matplotlib.pyplot as plt | |
| import numpy as np | |
| import pandas as pd | |
| from fastai.vision.all import * | |
| # Crear la clase | |
| class ADA_SKIN(object): | |
| # Inicializar el objeto | |
| def __init__(self, name="Wallaby", verbose=True, *args, **kwargs): | |
| super(ADA_SKIN, self).__init__(*args, **kwargs) | |
| self.author = "Jey" | |
| self.name = name | |
| if verbose: | |
| self._ph() | |
| self._pp("Hola desde la clase", str(self.__class__) + " Clase: " + str(self.__class__.__name__)) | |
| self._pp("Nombre del código", self.name) | |
| self._pp("Autor", self.author) | |
| self._ph() | |
| self.article = '<h3>Predice las siguientes patologias en piel</h3><ol>' | |
| self.article += '<li>Enfermedad de Bowen (AKIEC)</li>' | |
| self.article += '<li>Carcinoma de células basales</li>' | |
| self.article += '<li>Lesiones benignas similares a queratosis</li>' | |
| self.article += '<li>Dermatofibroma</li>' | |
| self.article += '<li>Melanoma</li>' | |
| self.article += '<li>Lunares melanocíticos</li>' | |
| self.article += '<li>Carcinoma de células escamosas</li>' | |
| self.article += '<li>Lesiones vasculares</li>' | |
| self.article += '<li>Benigno</li>' | |
| self.article += '<li></li></ol>' | |
| self.article += '<h3> Prueba Jey(2023)</h3><ul>' | |
| self.examples = ['akiec1.jpg','bcc1.jpg','bkl1.jpg','df1.jpg','mel1.jpg', | |
| 'nevi1.jpg','scc1.jpg','vl1.jpg','benign1.jpg','benign3.jpg'] | |
| self.title = "Predicción Cáncer de Piel prueba " | |
| return | |
| # Imprimir de manera legible el nombre y valor de una línea | |
| def _pp(self, a, b): | |
| print("%34s : %s" % (str(a), str(b))) | |
| return | |
| # Imprimir la línea de encabezado o pie de página | |
| def _ph(self): | |
| print("-" * 34, ":", "-" * 34) | |
| return | |
| def _predict_image(self, img, cat): | |
| # Convertir la imagen a PILImage | |
| img = PILImage.create(img) | |
| pred, idx, probs = learn.predict(img) | |
| return dict(zip(cat, map(float, probs))) | |
| def _predict_image2(self, img, cat): | |
| # Convertir la imagen a PILImage | |
| img = PILImage.create(img) | |
| pred, idx, probs = learn2.predict(img) | |
| return dict(zip(cat, map(float, probs))) | |
| def _draw_pred(self, df_pred, df2): | |
| fig, pic = plt.subplots(1, 2, figsize=(12, 6)) | |
| ti = df_pred["vocab"].head(3).values | |
| ti2 = df2["vocab"].head(2).values | |
| try: | |
| df_pred["pred"].head(3).plot(ax=pic[0], kind="pie", | |
| cmap="Set2", labels=ti, explode=(0.02, 0, 0), | |
| wedgeprops=dict(width=.4), | |
| normalize=False) | |
| df2["pred"].head(2).plot(ax=pic[1], kind="pie", | |
| colors=["cornflowerblue", "darkorange"], labels=ti2, explode=(0.02, 0), | |
| wedgeprops=dict(width=.4), | |
| normalize=False) | |
| except: | |
| df_pred["pred"].head(3).plot(ax=pic[0], kind="pie", | |
| cmap="Set2", labels=ti, explode=(0.02, 0, 0), | |
| wedgeprops=dict(width=.4)) | |
| df2["pred"].head(2).plot(ax=pic[1], kind="pie", | |
| colors=["cornflowerblue", "darkorange"], labels=ti2, explode=(0.02, 0), | |
| wedgeprops=dict(width=.4)) | |
| t = str(ti[0]) + ": " + str(np.round(df_pred.head(1).pred.values[0] * 100, 2)) + "% de predicción" | |
| pic[0].set_title(t, fontsize=14.0, fontweight="bold") | |
| pic[0].axis('off') | |
| pic[0].legend(ti, loc="lower right", title="Cáncer de Piel: ") | |
| k0 = np.round(df2.head(1).pred.values[0] * 100, 2) | |
| k1 = np.round(df2.tail(1).pred.values[0] * 100, 2) | |
| if k0 > k1: | |
| t2 = str(ti2[0]) + ": " + str(k0) + "% de predicción" | |
| else: | |
| t2 = str(ti2[1]) + ": " + str(k1) + "% de predicción" | |
| pic[1].set_title(t2, fontsize=14.0, fontweight="bold") | |
| pic[1].axis('off') | |
| pic[1].legend(ti2, loc="lower right", title="Predicción Cáncer de Piel:") | |
| fig.tight_layout() | |
| return fig | |
| def predict_donut(self, img): | |
| d = self._predict_image(img, self.categories) | |
| df = pd.DataFrame(d, index=[0]) | |
| df = df.transpose().reset_index() | |
| df.columns = ["vocab", "pred"] | |
| df.sort_values("pred", inplace=True, ascending=False, ignore_index=True) | |
| d2 = self._predict_image2(img, self.categories2) | |
| df2 = pd.DataFrame(d2, index=[0]) | |
| df2 = df2.transpose().reset_index() | |
| df2.columns = ["vocab", "pred"] | |
| fig = self._draw_pred(df, df2) | |
| return fig | |
| # Inicializar el objeto ADA_SKIN | |
| maxi = ADA_SKIN(verbose=False) | |
| # Cargar modelos | |
| learn = fastai.learner.load_learner('ada_learn_skin_norm2000.pkl') | |
| learn2 = fastai.learner.load_learner('ada_learn_malben.pkl') | |
| maxi.categories = learn.dls.vocab | |
| maxi.categories2 = learn2.dls.vocab | |
| # Crear la interfaz de Gradio | |
| hf_image = gr.Image() # Se elimina el argumento 'type' porque no es necesario | |
| hf_plot = gr.Plot() | |
| intf = gr.Interface(fn=maxi.predict_donut, | |
| inputs=hf_image, | |
| outputs=hf_plot, | |
| examples=maxi.examples, | |
| title=maxi.title, | |
| live=True, | |
| article=maxi.article) | |
| # Lanzar la interfaz | |
| intf.launch(share=True) | |