Search (2 results, page 1 of 1)

0.003754243 = product of:
  0.011262729 = sum of:
    0.011262729 = weight(_text_:a in 1868) [ClassicSimilarity], result of:
      0.011262729 = score(doc=1868,freq=16.0), product of:
        0.05209492 = queryWeight, product of:
          1.153047 = idf(docFreq=37942, maxDocs=44218)
          0.045180224 = queryNorm
        0.2161963 = fieldWeight in 1868, product of:
          4.0 = tf(freq=16.0), with freq of:
            16.0 = termFreq=16.0
          1.153047 = idf(docFreq=37942, maxDocs=44218)
          0.046875 = fieldNorm(doc=1868)
  0.33333334 = coord(1/3)

Abstract

We present a model that generates free-form natural language descriptions of image regions. Our model leverages datasets of images and their sentence descriptions to learn about the inter-modal correspondences between text and visual data. Our approach is based on a novel combination of Convolutional Neural Networks over image regions, bidirectional Recurrent Neural Networks over sentences, and a structured objective that aligns the two modalities through a multimodal embedding. We then describe a Recurrent Neural Network architecture that uses the inferred alignments to learn to generate novel descriptions of image regions. We demonstrate the effectiveness of our alignment model with ranking experiments on Flickr8K, Flickr30K and COCO datasets, where we substantially improve on the state of the art. We then show that the sentences created by our generative model outperform retrieval baselines on the three aforementioned datasets and a new dataset of region-level annotations.

Type

a

0.002473325 = product of:
  0.0074199745 = sum of:
    0.0074199745 = weight(_text_:a in 1865) [ClassicSimilarity], result of:
      0.0074199745 = score(doc=1865,freq=10.0), product of:
        0.05209492 = queryWeight, product of:
          1.153047 = idf(docFreq=37942, maxDocs=44218)
          0.045180224 = queryNorm
        0.14243183 = fieldWeight in 1865, product of:
          3.1622777 = tf(freq=10.0), with freq of:
            10.0 = termFreq=10.0
          1.153047 = idf(docFreq=37942, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1865)
  0.33333334 = coord(1/3)

Abstract

There's something magical about Recurrent Neural Networks (RNNs). I still remember when I trained my first recurrent network for Image Captioning. Within a few dozen minutes of training my first baby model (with rather arbitrarily-chosen hyperparameters) started to generate very nice looking descriptions of images that were on the edge of making sense. Sometimes the ratio of how simple your model is to the quality of the results you get out of it blows past your expectations, and this was one of those times. What made this result so shocking at the time was that the common wisdom was that RNNs were supposed to be difficult to train (with more experience I've in fact reached the opposite conclusion). Fast forward about a year: I'm training RNNs all the time and I've witnessed their power and robustness many times, and yet their magical outputs still find ways of amusing me. This post is about sharing some of that magic with you. By the way, together with this post I am also releasing code on Github (https://github.com/karpathy/char-rnn) that allows you to train character-level language models based on multi-layer LSTMs. You give it a large chunk of text and it will learn to generate text like it one character at a time. You can also use it to reproduce my experiments below. But we're getting ahead of ourselves; What are RNNs anyway?