{"componentChunkName":"component---src-templates-blog-post-js","path":"/blog/semantic-segmentation-with-tensorflow-keras/","result":{"data":{"site":{"siteMetadata":{"title":"No Frills News"}},"contentfulNfnPost":{"postTitle":"Semantic Segmentation with TensorFlow Keras","slug":"semantic-segmentation-with-tensorflow-keras","createdLocal":"2021-05-15 14:30:51.892788","publishDate":"2021-05-15 07:30:00+00:00","feedName":"Image Recognition","sourceUrl":{"sourceUrl":"https://analyticsindiamag.com/semantic-segmentation-using-tensorflow-keras/"},"postSummary":{"childMarkdownRemark":{"html":"<p>In this article, we discuss semantic segmentation using TensorFlow Keras.\nAmong FCNNs, U-Net is one of the successful architectures acclaimed for its performance in Medical Image Segmentation.\ndef brightness(img, mask): img = tf.image.adjust<em>brightness(img, 0.1) return img, mask def gamma(img, mask): img = tf.image.adjust</em>gamma(img, 0.1) return img, mask def hue(img, mask): img = tf.image.adjust<em>hue(img, -0.1) return img, mask def crop(img, mask): img = tf.image.central</em>crop(img, 0.7) img = tf.image.resize(img, (128,128)) mask = tf.image.central<em>crop(mask, 0.7) mask = tf.image.resize(mask, (128,128)) mask = tf.cast(mask, tf.uint8) return img, mask def flip</em>hori(img, mask): img = tf.image.flip<em>left</em>right(img) mask = tf.image.flip<em>left</em>right(mask) return img, mask def flip<em>vert(img, mask): img = tf.image.flip</em>up<em>down(img) mask = tf.image.flip</em>up<em>down(mask) return img, mask def rotate(img, mask): img = tf.image.rot90(img) mask = tf.image.rot90(mask) return img, maskApply augmentation to the data with the above functions.\ntrain = tf.data.Dataset.zip((train</em>X, train<em>y)) val = tf.data.Dataset.zip((val</em>X, val<em>y)) # perform augmentation on train data only a = train.map(brightness) b = train.map(gamma) c = train.map(hue) d = train.map(crop) e = train.map(flip</em>hori) f = train.map(flip_vert) g = train.map(rotate) train = train.concatenate(a) train = train.concatenate(b) train = train.concatenate(c) train = train.concatenate(d) train = train.concatenate(e) train = train.concatenate(f) train = train.concatenate(g)Prepare data batches.\nWrapping UpIn this article, we have discussed semantic segmentation using TensorFlow Keras.</p>"}}}},"pageContext":{"slug":"semantic-segmentation-with-tensorflow-keras"}}}