Understanding Low- and High-Level Contributions to Fixation Prediction
01 Oct 2017-pp 4799-4808
TL;DR: Comparing different features within the same powerful readout architecture allows to better understand the relevance of low- versus high-level features in predicting fixation locations, while simultaneously achieving state-of-the-art saliency prediction.
Abstract: Understanding where people look in images is an important problem in computer vision. Despite significant research, it remains unclear to what extent human fixations can be predicted by low-level (contrast) compared to highlevel (presence of objects) image features. Here we address this problem by introducing two novel models that use different feature spaces but the same readout architecture. The first model predicts human fixations based on deep neural network features trained on object recognition. This model sets a new state-of-the art in fixation prediction by achieving top performance in area under the curve metrics on the MIT300 hold-out benchmark (AUC = 88%, sAUC = 77%, NSS = 2.34). The second model uses purely low-level (isotropic contrast) features. This model achieves better performance than all models not using features pretrained on object recognition, making it a strong baseline to assess the utility of high-level features. We then evaluate and visualize which fixations are better explained by lowlevel compared to high-level image features. Surprisingly we find that a substantial proportion of fixations are better explained by the simple low-level model than the stateof- the-art model. Comparing different features within the same powerful readout architecture allows us to better understand the relevance of low- versus high-level features in predicting fixation locations, while simultaneously achieving state-of-the-art saliency prediction.
Citations
More filters
TL;DR: This protocol describes how to use an open-source toolbox, DeepLabCut, to train a deep neural network to precisely track user-defined features with limited training data, which allows noninvasive behavioral tracking of movement.
Abstract: Noninvasive behavioral tracking of animals during experiments is critical to many scientific pursuits. Extracting the poses of animals without using markers is often essential to measuring behavioral effects in biomechanics, genetics, ethology, and neuroscience. However, extracting detailed poses without markers in dynamically changing backgrounds has been challenging. We recently introduced an open-source toolbox called DeepLabCut that builds on a state-of-the-art human pose-estimation algorithm to allow a user to train a deep neural network with limited training data to precisely track user-defined features that match human labeling accuracy. Here, we provide an updated toolbox, developed as a Python package, that includes new features such as graphical user interfaces (GUIs), performance improvements, and active-learning-based network refinement. We provide a step-by-step procedure for using DeepLabCut that guides the user in creating a tailored, reusable analysis pipeline with a graphical processing unit (GPU) in 1–12 h (depending on frame size). Additionally, we provide Docker environments and Jupyter Notebooks that can be run on cloud resources such as Google Colaboratory. This protocol describes how to use an open-source toolbox, DeepLabCut, to train a deep neural network to precisely track user-defined features with limited training data. This allows noninvasive behavioral tracking of movement.
674 citations
TL;DR: Zhang et al. as mentioned in this paper proposed a convolutional long short-term memory (LSTM) network to iteratively refine the predicted saliency map by focusing on the most salient regions of the input image.
Abstract: Data-driven saliency has recently gained a lot of attention thanks to the use of convolutional neural networks for predicting gaze fixations. In this paper, we go beyond standard approaches to saliency prediction, in which gaze maps are computed with a feed-forward network, and present a novel model which can predict accurate saliency maps by incorporating neural attentive mechanisms. The core of our solution is a convolutional long short-term memory that focuses on the most salient regions of the input image to iteratively refine the predicted saliency map. In addition, to tackle the center bias typical of human eye fixations, our model can learn a set of prior maps generated with Gaussian functions. We show, through an extensive evaluation, that the proposed architecture outperforms the current state-of-the-art on public saliency prediction datasets. We further study the contribution of each key component to demonstrate their robustness on different scenarios.
503 citations
TL;DR: A database collecting subjects’ HM in panoramic video sequences is established, and it is found that deep reinforcement learning (DRL) can be applied to predict HM positions, via maximizing the reward of imitating human HM scanpaths through the agent's actions.
Abstract: Panoramic video provides immersive and interactive experience by enabling humans to control the field of view (FoV) through head movement (HM). Thus, HM plays a key role in modeling human attention on panoramic video. This paper establishes a database collecting subjects’ HM in panoramic video sequences. From this database, we find that the HM data are highly consistent across subjects. Furthermore, we find that deep reinforcement learning (DRL) can be applied to predict HM positions, via maximizing the reward of imitating human HM scanpaths through the agent's actions. Based on our findings, we propose a DRL-based HM prediction (DHP) approach with offline and online versions, called offline-DHP and online-DHP. In offline-DHP, multiple DRL workflows are run to determine potential HM positions at each panoramic frame. Then, a heat map of the potential HM positions, named the HM map, is generated as the output of offline-DHP. In online-DHP, the next HM position of one subject is estimated given the currently observed HM position, which is achieved by developing a DRL algorithm upon the learned offline-DHP model. Finally, the experiments validate that our approach is effective in both offline and online prediction of HM positions for panoramic video, and that the learned offline-DHP model can improve the performance of online-DHP.
166 citations
Cites background from "Understanding Low- and High-Level C..."
...In [35], a readout architecturewas proposed to predict human attention on static images, in which both DNN features and low-level (isotropic contrast) features are considered....
[...]
...Benefiting from the most recent success of deep learning, deep neural networks (DNNs) [14], [15], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36] have also been developed to detect 2D video saliency, rather than exploring the HVSrelated features as in heuristic saliency detection approaches....
[...]
TL;DR: A large number of image and video saliency models are reviewed and compared over two image benchmarks and two large scale video datasets and factors that contribute to the gap between models and humans are identified.
Abstract: Visual saliency models have enjoyed a big leap in performance in recent years, thanks to advances in deep learning and large scale annotated data. Despite enormous effort and huge breakthroughs, however, models still fall short in reaching human-level accuracy. In this work, I explore the landscape of the field emphasizing on new deep saliency models, benchmarks, and datasets. A large number of image and video saliency models are reviewed and compared over two image benchmarks and two large scale video datasets. Further, I identify factors that contribute to the gap between models and humans and discuss the remaining issues that need to be addressed to build the next generation of more powerful saliency models. Some specific questions that are addressed include: in what ways current models fail, how to remedy them, what can be learned from cognitive studies of attention, how explicit saliency judgments relate to fixations, how to conduct fair model comparison, and what are the emerging applications of saliency models.
133 citations
Cites background or methods from "Understanding Low- and High-Level C..."
...[65] conclude that no matter which score is used, there still might be images for which a weaker model performs better than the best model....
[...]
...B) Right: Performances of DeepGaze II and ICF models over the MIT1003 dataset [65]....
[...]
...More recently, they introduced the DeepGaze II model [65] built upon DeepGaze I....
[...]
TL;DR: This work proposes an end-to-end dilated inception network (DINet) for visual saliency prediction that captures multi-scale contextual features effectively with very limited extra parameters and improves the performance of the saliency model by using a set of linear normalization-based probability distribution distance metrics as loss functions.
Abstract: Recently, with the advent of deep convolutional neural networks (DCNN), the improvements in visual saliency prediction research are impressive. One possible direction to approach the next improvement is to fully characterize the multi-scale saliency-influential factors with a computationally-friendly module in DCNN architectures. In this work, we propose an end-to-end dilated inception network (DINet) for visual saliency prediction. It captures multi-scale contextual features effectively with very limited extra parameters. Instead of utilizing parallel standard convolutions with different kernel sizes as the existing inception module, our proposed dilated inception module (DIM) uses parallel dilated convolutions with different dilation rates which can significantly reduce the computation load while enriching the diversity of receptive fields in feature maps. Moreover, the performance of our saliency model is further improved by using a set of linear normalization-based probability distribution distance metrics as loss functions. As such, we can formulate saliency prediction as a global probability distribution prediction task for better saliency inference instead of a pixel-wise regression problem. Experimental results on several challenging saliency benchmark datasets demonstrate that our DINet with proposed loss functions can achieve state-of-the-art performance with shorter inference time.
101 citations
Cites background from "Understanding Low- and High-Level C..."
...The DeepGazeII [38] model get the best sAUC score and relatively lower scores on other...
[...]
...The DeepGazeII [38] model get the best sAUC score and relatively lower scores on other metrics....
[...]
References
More filters
Proceedings Article•
03 Dec 2012TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
Abstract: We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overriding in the fully-connected layers we employed a recently-developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry.
73,978 citations
Proceedings Article•
01 Jan 2015TL;DR: In this paper, the authors investigated the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting and showed that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 layers.
Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.
49,914 citations
20 Jun 2009
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
Abstract: The explosion of image data on the Internet has the potential to foster more sophisticated and robust models and algorithms to index, retrieve, organize and interact with images and multimedia data. But exactly how such data can be harnessed and organized remains a critical problem. We introduce here a new database called “ImageNet”, a large-scale ontology of images built upon the backbone of the WordNet structure. ImageNet aims to populate the majority of the 80,000 synsets of WordNet with an average of 500-1000 clean and full resolution images. This will result in tens of millions of annotated images organized by the semantic hierarchy of WordNet. This paper offers a detailed analysis of ImageNet in its current state: 12 subtrees with 5247 synsets and 3.2 million images in total. We show that ImageNet is much larger in scale and diversity and much more accurate than the current image datasets. Constructing such a large-scale database is a challenging task. We describe the data collection scheme with Amazon Mechanical Turk. Lastly, we illustrate the usefulness of ImageNet through three simple applications in object recognition, image classification and automatic object clustering. We hope that the scale, accuracy, diversity and hierarchical structure of ImageNet can offer unparalleled opportunities to researchers in the computer vision community and beyond.
49,639 citations
"Understanding Low- and High-Level C..." refers methods in this paper
...The models make use of convolutional filters that have been learned on other tasks, most notably object recognition in the ImageNet dataset [10]....
[...]
...Subsequently, the DeepGaze I model showed that DNN features trained on object recognition (AlexNet [29] trained on the ImageNet dataset [10]) could significantly outperform training from scratch [32]....
[...]
07 Jun 2015
TL;DR: Inception as mentioned in this paper is a deep convolutional neural network architecture that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14).
Abstract: We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.
40,257 citations
Posted Content•
TL;DR: Caffe as discussed by the authors is a BSD-licensed C++ library with Python and MATLAB bindings for training and deploying general-purpose convolutional neural networks and other deep models efficiently on commodity architectures.
Abstract: Caffe provides multimedia scientists and practitioners with a clean and modifiable framework for state-of-the-art deep learning algorithms and a collection of reference models. The framework is a BSD-licensed C++ library with Python and MATLAB bindings for training and deploying general-purpose convolutional neural networks and other deep models efficiently on commodity architectures. Caffe fits industry and internet-scale media needs by CUDA GPU computation, processing over 40 million images a day on a single K40 or Titan GPU ($\approx$ 2.5 ms per image). By separating model representation from actual implementation, Caffe allows experimentation and seamless switching among platforms for ease of development and deployment from prototyping machines to cloud environments. Caffe is maintained and developed by the Berkeley Vision and Learning Center (BVLC) with the help of an active community of contributors on GitHub. It powers ongoing research projects, large-scale industrial applications, and startup prototypes in vision, speech, and multimedia.
12,531 citations