Higher Attention Scores Research Articles

Transformer-Based Visual Object Tracking with Global Feature Enhancement

With the rise of general models, transformers have been adopted in visual object tracking algorithms as feature fusion networks. In these trackers, self-attention is used for global feature enhancement. Cross-attention is applied to fuse the features of the template and the search regions to capture the global information of the object. However, studies have found that the feature information fused by cross-attention does not pay enough attention to the object region. In order to enhance cross-attention for the object region, an enhanced cross-attention (ECA) module is proposed for global feature enhancement. By calculating the average attention score for each position in the fused feature sequence and assigning higher weights to the positions with higher attention scores, the proposed ECA module can improve the feature information in the object region and further enhance the matching accuracy. In addition, to reduce the computational complexity of self-attention, orthogonal random features are introduced to implement a fast attention operation. This decomposes the attention matrix into the product of a random non-linear function between the original query and key. This module can reduce the spatial complexity and improve the inference speed by avoiding the explicit construction of a quadratic attention matrix. Finally, a tracking method named GFETrack is proposed, which comprises a Siamese backbone network and an enhanced attention mechanism. Experimental results show that the proposed GFETrack achieves competitive results on four challenging datasets.

Prenatal Nicotine or Cannabis Exposure and Offspring Neurobehavioral Outcomes.

To study the association between nicotine or cannabis metabolite presence in maternal urine and child neurodevelopmental outcomes. We conducted a secondary analysis of two parallel multicenter randomized controlled trials of treatment for hypothyroxinemia or subclinical hypothyroidism among pregnant individuals enrolled at 8-20 weeks of gestation. All maternal-child dyads with a maternal urine sample at enrollment and child neurodevelopmental testing were included (N=1,197). Exposure was urine samples positive for nicotine (cotinine) or cannabis 11-nor-9-carboxy-delta-9-tetrahydrocannabinol [THC-COOH]) or both metabolites. Primary outcome was child IQ at 60 months. Secondary outcomes included cognitive, motor and language, attention, behavioral and social competency, and differential skills assessments at 12, 24, 36, and 48 months. Quantile regression analysis was performed with confounder adjustment. Of 1,197 pregnant individuals, 99 (8.3%) had positive cotinine samples and 47 (3.9%) had positive THC-COOH samples; 33 (2.8%) were positive for both. Groups differed in self-reported race and ethnicity, education, marital status, insurance, and thyroid status. Median IQ was similar between cotinine-exposed and -unexposed children (90 vs 95, adjusted difference in medians -2.47, 95% CI -6.22 to 1.29) and THC-COOH-exposed and -unexposed children (89 vs 95, adjusted difference in medians -1.35, 95% CI -7.76 to 5.05). In secondary outcome analysis, children with THC-COOH exposure compared with those unexposed had higher attention scores at 48 months of age (57 vs 49, adjusted difference in medians 6.0, 95% CI 1.11-10.89). Neither prenatal nicotine nor cannabis exposure was associated with a difference in IQ. Cannabis exposure was associated with worse attention scores in early childhood. Longitudinal studies assessing associations between child neurodevelopmental outcomes and prenatal nicotine and cannabis exposure with a focus on timing and quantity of exposure are needed. ClinicalTrials.gov, NCT00388297.

Do authors of research funded by the Canadian Institutes of Health Research comply with its open access mandate?: A meta-epidemiologic study

BackgroundSince 2008, the Canadian Institutes of Health Research (CIHR) has mandated that studies it funds either in whole or in part are required to publish their results as open access (OA) within 12 months of publication using either online repositories and/or OA journals. Yet, there is evidence that authors are poorly compliant with this mandate. Specifically, there has been an apparent decrease in OA publication after 2015, which coincides with a change in the OA policy during the same year. One particular policy change that may have contributed to this decline was lifting the requirement that authors deposit their article in an OA repository immediately upon publication. We investigated the proportion of OA compliance of CIHR-funded studies in the period before and after the policy change of 2015 with manual confirmation of both CIHR funding and OA status.Methods and findingsWe identified CIHR-funded studies published between the years 2014 to 2017 using a comprehensive search in the Web of Science (WoS). We took a stratified random sample from all four years (i.e. 2014 to 2017), with 250 studies from each year. Two authors independently reviewed the final full-text publications retrieved from the journal web page to determine to confirm CIHR funding, as indicated in the acknowledgements or elsewhere in the paper. For each study, we also collected bibliometric data that included citation count and Altmetric attention score Statistical analyses were conducted using two-tailed Fisher’s exact test with relative risk (RR). Among the 851 receiving CIHR funding published from 2014 to 2017, the percentage of CIHR-funded studies published as OA significantly decreased from 79.6% in 2014 to 70.3% in 2017 (RR = 0.88, 95% CI: 0.79–0.99, P = 0.028). When considering all four years, there was no significant difference in the percentage of CIHR-funded studies published as OA in both 2014 and 2015 compared to both 2016 and 2017 (RR = 0.97, 95% CI: 0.90–1.05, P = 0.493). Additionally, OA publications had significantly higher citation count (both in year of publication and in total) and higher attention scores (P<0.05).ConclusionsOverall, we found that there was a significant decrease in the proportion of CIHR funded studies published as OA from 2014 compared to 2017, though this difference did not persist when comparing both 2014–2015 to 2016–2017. The primary limitation was the reliance of self-reported data from authors on CIHR funding status. We posit that this decrease may be attributable to CIHR’s OA policy change in 2015. Further exploration is warranted to both validate these studies using a larger dataset and, if valid, investigate the effects of potential interventions to improve the OA compliance, such as use of a CIHR publication database, and reinstatement of a policy for authors to immediately submit their findings to OA repositories upon publication.

Online processing of sentences containing noun modification in young children with high-functioning autism.

There is variability in the language of children with autism, even those who are high functioning. However, little is known about how they process language structures in real time, including how they handle potential ambiguity, and whether they follow referential constraints. Previous research with older autism spectrum disorder (ASD) participants has shown that these individuals can use context to access rapidly the meaning of ambiguous words. The severity of autism has also been shown to influence the speed in which children with ASD access lexical information. To understand more about how children with ASD process language in real time (i.e., as it unfolds). The focus was the integration of information and use of referential constraints to identify a referent named in a sentence. We used an eye-tracking task to compare performance between young, high-functioning children with autism (HFA) and children with typical development (TD). A large sample of 5-9-year-old children (mean age = 6;8 years), 48 with HFA and 56 with TD participated; all were attending mainstream schools. For each item participants were shown a display of four images that differed in two dimensions. Each sentence contained an adjective and noun that restricted the choice from four to two (the target and competitor), followed by a prepositional phrase (e.g., the blue square with dots); this added modifying information to provide a unique description of the target. We calculated looking time at the target, the competitor and the two distractors for each 200 ms time interval as children processed the sentence and looked at the display. Generalized estimating equations were used to carry out repeated-measures analyses on the proportion of looking time to target and competitor and time to fixate to target. Children in both groups (HFA and TD) looked at the target and competitor more than at the distractors following the adjective and noun and following the modifying information in the prepositional phrase more at the target. However, the HFA group was significantly slower in both phases and looked proportionally less at the target. Across the sample, IQ and language did not affect the results; however, age and attention had an impact. The older children showed an advantage in processing the information as did the children with higher attention scores. The HFA group took longer than the TD group to integrate the disambiguating information provided in the course of processing a sentence and integrate it with the visual information, indicating that for the ASD group incremental processing was not as advanced as for children with ASD, and they were less sensitive to referential conventions. Training for young children with ASD on the use of referential conventions and available contextual clues may be of benefit to them in understanding the language they hear.

Heart rate variability biofeedback, executive functioning and chronic brain injury

Primary objective: To determine if individuals with brain injury can modify heart rate variability (HRV) through biofeedback and, if so, enhance its pattern to improve emotional regulation and problem-solving ability.Design: A quasi-experimental design with repeated measures was employed. Thirteen individuals aged 23–63 years with severe brain injury (13–40 years post-onset) participating in a community-based programme were enrolled.Main outcomes: Response-to-treatment was measured with HRV indices, Behavior Rating Inventory of Executive Function (BRIEF-A-Informant) and attention/problem-solving tests.Results: At post-treatment, HRV indices (Low Frequency/High Frequency [LF/HF] and coherence ratio) increased significantly. Increased LF/HF values during the second-half of a 10-minute session were associated with higher attention scores. Participants who scored better (by scoring lower) in informant ratings at pre-treatment had highest HRV scores at post-treatment. Accordingly, at post-treatment, families’ ratings of participants’ emotional control correlated with HRV indices; staffs’ ratings of participants’ working memory correlated with participants’ HRV indices. Self-ratings of the BRIEF-A Task Monitoring scale at post-treatment correlated with family ratings at pre-treatment and post-treatment.Conclusions: Results demonstrate an association between regulation of emotions/cognition and HRV training. Individuals with severe, chronic brain injury can modify HRV through biofeedback. Future research should evaluate the efficacy of this approach for modifying behavioural problems.