skip to main content

Attention:

The NSF Public Access Repository (PAR) system and access will be unavailable from 8:00 PM ET on Friday, March 21 until 8:00 AM ET on Saturday, March 22 due to maintenance. We apologize for the inconvenience.


Title: COCO-Search18 fixation dataset for predicting goal-directed attention control
Abstract

Attention control is a basic behavioral process that has been studied for decades. The currently best models of attention control are deep networks trained on free-viewing behavior to predict bottom-up attention control – saliency. We introduce COCO-Search18, the first dataset of laboratory-qualitygoal-directed behaviorlarge enough to train deep-network models. We collected eye-movement behavior from 10 people searching for each of 18 target-object categories in 6202 natural-scene images, yielding$$\sim$$300,000 search fixations. We thoroughly characterize COCO-Search18, and benchmark it using three machine-learning methods: a ResNet50 object detector, a ResNet50 trained on fixation-density maps, and an inverse-reinforcement-learning model trained on behavioral search scanpaths. Models were also trained/tested on images transformed to approximate a foveated retina, a fundamental biological constraint. These models, each having a different reliance on behavioral training, collectively comprise the new state-of-the-art in predicting goal-directed search fixations. Our expectation is that future work using COCO-Search18 will far surpass these initial efforts, finding applications in domains ranging from human-computer interactive systems that can anticipate a person’s intent and render assistance to the potentially early identification of attention-related clinical disorders (ADHD, PTSD, phobia) based on deviation from neurotypical fixation behavior.

 
more » « less
Award ID(s):
1763981 1650499
PAR ID:
10223276
Author(s) / Creator(s):
; ; ; ; ;
Publisher / Repository:
Nature Publishing Group
Date Published:
Journal Name:
Scientific Reports
Volume:
11
Issue:
1
ISSN:
2045-2322
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Human gaze behavior prediction is important for behavioral vision and for computer vision applications. Most models mainly focus on predicting free-viewing behavior using saliency maps, but do not generalize to goal-directed behavior, such as when a person searches for a visual target object. We propose the first inverse reinforcement learning (IRL) model to learn the internal reward function and policy used by humans during visual search. We modeled the viewer’s internal belief states as dynamic contextual belief maps of object locations. These maps were learned and then used to predict behavioral scanpaths for multiple target categories. To train and evaluate our IRL model we created COCO-Search18, which is now the largest dataset of highquality search fixations in existence. COCO-Search18 has 10 participants searching for each of 18 target-object categories in 6202 images, making about 300,000 goal-directed fixations. When trained and evaluated on COCO-Search18, the IRL model outperformed baseline models in predicting search fixation scanpaths, both in terms of similarity to human search behavior and search efficiency. Finally, reward maps recovered by the IRL model reveal distinctive targetdependent patterns of object prioritization, which we interpret as a learned object context. 
    more » « less
  2. Abstract

    This paper presents a search for dark matter,$$\chi $$χ, using events with a single top quark and an energeticWboson. The analysis is based on proton–proton collision data collected with the ATLAS experiment at$$\sqrt{s}=$$s=13 TeV during LHC Run 2 (2015–2018), corresponding to an integrated luminosity of 139 fb$$^{-1}$$-1. The search considers final states with zero or one charged lepton (electron or muon), at least oneb-jet and large missing transverse momentum. In addition, a result from a previous search considering two-charged-lepton final states is included in the interpretation of the results. The data are found to be in good agreement with the Standard Model predictions and the results are interpreted in terms of 95% confidence-level exclusion limits in the context of a class of dark matter models involving an extended two-Higgs-doublet sector together with a pseudoscalar mediator particle. The search is particularly sensitive to on-shell production of the charged Higgs boson state,$$H^{\pm }$$H±, arising from the two-Higgs-doublet mixing, and its semi-invisible decays via the mediator particle,a:$$H^{\pm } \rightarrow W^\pm a (\rightarrow \chi \chi )$$H±W±a(χχ). Signal models with$$H^{\pm }$$H±masses up to 1.5 TeV andamasses up to 350 GeV are excluded assuming a$$\tan \beta $$tanβvalue of 1. For masses ofaof 150 (250) GeV,$$\tan \beta $$tanβvalues up to 2 are excluded for$$H^{\pm }$$H±masses between 200 (400) GeV and 1.5 TeV. Signals with$$\tan \beta $$tanβvalues between 20 and 30 are excluded for$$H^{\pm }$$H±masses between 500 and 800 GeV.

     
    more » « less
  3. A<sc>bstract</sc>

    A search for supersymmetry is presented in events with a single charged lepton, electron or muon, and multiple hadronic jets. The data correspond to an integrated luminosity of 138 fb1of proton-proton collisions at a center-of-mass energy of 13 TeV, recorded by the CMS experiment at the CERN LHC. The search targets gluino pair production, where the gluinos decay into final states with the lightest supersymmetric particle (LSP) and either a top quark-antiquark ($$ \textrm{t}\overline{\textrm{t}} $$tt¯) pair, or a light-flavor quark-antiquark ($$ \textrm{q}\overline{\textrm{q}} $$qq¯) pair and a virtual or on-shell W boson. The main backgrounds,$$ \textrm{t}\overline{\textrm{t}} $$tt¯pair and W+jets production, are suppressed by requirements on the azimuthal angle between the momenta of the lepton and of its reconstructed parent W boson candidate, and by top quark and W boson identification based on a machine-learning technique. The number of observed events is consistent with the expectations from standard model processes. Limits are evaluated on supersymmetric particle masses in the context of two simplified models of gluino pair production. Exclusions for gluino masses reach up to 2120 (2050) GeV at 95% confidence level for a model with gluino decay to a$$ \textrm{t}\overline{\textrm{t}} $$tt¯pair (a$$ \textrm{q}\overline{\textrm{q}} $$qq¯pair and a W boson) and the LSP. For the same models, limits on the mass of the LSP reach up to 1250 (1070) GeV.

     
    more » « less
  4. Abstract

    A search for supersymmetry involving the pair production of gluinos decaying via off-shell third-generation squarks into the lightest neutralino$$(\tilde{\chi }^0_1)$$(χ~10)is reported. It exploits LHC proton–proton collision data at a centre-of-mass energy$$\sqrt{s} = 13$$s=13TeV with an integrated luminosity of 139 fb$$^{-1}$$-1collected with the ATLAS detector from 2015 to 2018. The search uses events containing large missing transverse momentum, up to one electron or muon, and several energetic jets, at least three of which must be identified as containingb-hadrons. Both a simple kinematic event selection and an event selection based upon a deep neural-network are used. No significant excess above the predicted background is found. In simplified models involving the pair production of gluinos that decay via off-shell top (bottom) squarks, gluino masses less than 2.44 TeV (2.35 TeV) are excluded at 95% CL for a massless$$\tilde{\chi }^0_1.$$χ~10.Limits are also set on the gluino mass in models with variable branching ratios for gluino decays to$$b\bar{b}\tilde{\chi }^0_1,$$bb¯χ~10,$$t\bar{t}\tilde{\chi }^0_1$$tt¯χ~10and$$t\bar{b}\tilde{\chi }^-_1/\bar{t}b\tilde{\chi }^+_1.$$tb¯χ~1-/t¯bχ~1+.

     
    more » « less
  5. Abstract

    The need to determine the structure of a graph arises in many applications. This paper studies directed graphs and defines the notions of$$\ell$$-chained and$$\{\ell ,k\}$${,k}-chained directed graphs. These notions reveal structural properties of directed graphs that shed light on how the nodes of the graph are connected. Applications include city planning, information transmission, and disease propagation. We also discuss the notion of in-center and out-center vertices of a directed graph, which are vertices at the center of the graph. Computed examples provide illustrations, among which is the investigation of a bus network for a city.

     
    more » « less