Tipping the Scales Toward Single Modality

2022 ◽  
Vol 112 (2) ◽  
pp. 279-280
Author(s):  
Stephanie K. Schaub ◽  
Catherine M. Albert
Keyword(s):  
Single Modality

scholarly journals Bottom-up and Layerwise Domain Adaptation for Pedestrian Detection in Thermal Images

2021 ◽  
Vol 17 (1) ◽  
pp. 1-19
Author(s):  
My Kieu ◽  
Andrew D. Bagdanov ◽  
Marco Bertini
Keyword(s):  
Domain Adaptation ◽  
State Of The Art ◽  
Pedestrian Detection ◽  
Challenging Problem ◽  
Top Down ◽  
Bottom Up ◽  
Security Applications ◽  
Lighting Conditions ◽  
Initial Layers ◽  
Single Modality

Pedestrian detection is a canonical problem for safety and security applications, and it remains a challenging problem due to the highly variable lighting conditions in which pedestrians must be detected. This article investigates several domain adaptation approaches to adapt RGB-trained detectors to the thermal domain. Building on our earlier work on domain adaptation for privacy-preserving pedestrian detection, we conducted an extensive experimental evaluation comparing top-down and bottom-up domain adaptation and also propose two new bottom-up domain adaptation strategies. For top-down domain adaptation, we leverage a detector pre-trained on RGB imagery and efficiently adapt it to perform pedestrian detection in the thermal domain. Our bottom-up domain adaptation approaches include two steps: first, training an adapter segment corresponding to initial layers of the RGB-trained detector adapts to the new input distribution; then, we reconnect the adapter segment to the original RGB-trained detector for final adaptation with a top-down loss. To the best of our knowledge, our bottom-up domain adaptation approaches outperform the best-performing single-modality pedestrian detection results on KAIST and outperform the state of the art on FLIR.

Machine Learning Multi-Modality Fusion Approaches Outperform Single-Modality & Traditional Approaches

2021 ◽  
Author(s):  
Denis Garagic ◽  
Daniel Pelgrift ◽  
Jacob Peskoe ◽  
Ronald D. Hagan ◽  
Peter Zulch ◽  
...  
Keyword(s):  
Machine Learning ◽  
Traditional Approaches ◽  
Single Modality

Challenges in the Management of Ruptured and Unruptured Brainstem Arteriovenous Malformations: Outcome After Conservative, Single-Modality, or Multimodality Treatments

Neurosurgery ◽  
2011 ◽  
Vol 70 (1) ◽  
pp. 155-161 ◽  
Author(s):  
Laurent Thines ◽  
Amir R. Dehdashti ◽  
Leodante da Costa ◽  
Michael Tymianski ◽  
Karel G. ter Brugge ◽  
...  
Keyword(s):  
Scale Score ◽  
Odds Ratio ◽  
Arteriovenous Malformations ◽  
Negative Impact ◽  
Multimodality Treatment ◽  
Treatment Modalities ◽  
Modified Rankin Scale ◽  
Mortality And Morbidity ◽  
Cure Rates ◽  
Single Modality

Abstract BACKGROUND Brainstem arteriovenous malformations are challenging lesions, and benefits of treatment are uncertain. OBJECTIVE To study the clinical course of Brainstem arteriovenous malformations and the influence of treatments on outcome. METHODS We reviewed a prospective series of 31 brainstem arteriovenous malformations. Demographic, morphological, and clinical characteristics were recorded. Factors determining initial and final outcomes (modified Rankin Scale), results of treatments (cure rates, complications), and disease course were analyzed. RESULTS Brainstem arteriovenous malformations were symptomatic and bled in 93% and 61% of cases, respectively. Examination was abnormal and initial modified Rankin Scale score was > 3 in 71% and 86% of patients, respectively. The average follow-up time was 6.2 years, and 26% of patients rebled (5.9 %/y). Treatment modalities included conservative, radiosurgical, endovascular, surgical, and multimodality treatment in 13%, 58%, 35%, 16%, and 26% of cases, respectively. The obliteration rate was 60% overall and 39% after radiosurgery, 40% after embolization, and 75% after microsurgery, with respective complication-free cure rates of 71%, 50%, and 0%. Overall procedural mortality and morbidity were 2.3% and 18.6%, respectively. Final modified Rankin Scale score was > 3 in 77% of cases. Neurological deterioration (35%) was related to treatment complications in 74% of cases with a negative impact of surgery (P = .04), palliative embolization (odds ratio = 16), and multimodality treatments (odds ratio = 24). Radiosurgery was inversely associated with worsening (odds ratio = 0.06). CONCLUSION Brainstem arteriovenous malformations require individualized treatment decisions. Single-modality treatments with a reasonable chance of complete cure and low complication rate (such as radiosurgery) should be favored.

scholarly journals Multimodal Sensor Data Fusion for Activity Recognition Using Filtered Classifier

Proceedings ◽  
2018 ◽  
Vol 2 (19) ◽  
pp. 1262 ◽  
Author(s):  
Muhammad Razzaq ◽  
Ian Cleland ◽  
Chris Nugent ◽  
Sungyoung Lee
Keyword(s):  
Activity Recognition ◽  
Confusion Matrix ◽  
Machine Learning Algorithms ◽  
Training Data ◽  
Sensor Data ◽  
Test Dataset ◽  
Sensory Data ◽  
Multimodal Sensors ◽  
Imbalanced Class Distribution ◽  
Single Modality

Activity recognition (AR) is a subtask in pervasive computing and context-aware systems, which presents the physical state of human in real-time. These systems offer a new dimension to the widely spread applications by fusing recognized activities obtained from the raw sensory data generated by the obtrusive as well as unobtrusive revolutionary digital technologies. In recent years, an exponential growth has been observed for AR technologies and much literature exists focusing on applying machine learning algorithms on obtrusive single modality sensor devices. However, University of Jaén Ambient Intelligence (UJAmI), a Smart Lab in Spain has initiated a 1st UCAmI Cup challenge by sharing aforementioned varieties of the sensory data in order to recognize the human activities in the smart environment. This paper presents the fusion, both at the feature level and decision level for multimodal sensors by preprocessing and predicting the activities within the context of training and test datasets. Though it achieves 94% accuracy for training data and 47% accuracy for test data. However, this study further evaluates post-confusion matrix also and draws a conclusion for various discrepancies such as imbalanced class distribution within the training and test dataset. Additionally, this study also highlights challenges associated with the datasets for which, could improve further analysis.

The impact of Radiotherapy on Swallowing and Speech in Patients who Undergo total Laryngectomy

2008 ◽  
Vol 139 (6) ◽  
pp. 792-797 ◽  
Author(s):  
Carmen de Casso ◽  
Nicholas J. Slevin ◽  
Jarrod J. Homer
Keyword(s):  
Total Laryngectomy ◽  
Chart Review ◽  
Detrimental Effect ◽  
Definitive Treatment ◽  
Soft Diet ◽  
Carcinoma Of The Larynx ◽  
Voice Development ◽  
The Impact ◽  
Single Modality

Objectives Quality of life studies have shown no detrimental effect with radiotherapy (RT) in patients who have a total laryngectomy. We wished to determine the effect of RT (initial or postoperative) specifically on the swallowing and voice function in patients treated by total laryngectomy (TL) for carcinoma of the larynx. Design Multicenter chart review. Setting Multicenter study in the Greater Manchester and Lancashire area. Participants A total of 121 postlaryngectomy patients all of whom had completed definitive treatment at least 6 months before this study. Twenty-six patients had total laryngectomy as a single modality treatment and 95 had total laryngectomy and radiotherapy. Main Outcome Measures Swallowing (solid food, soft diet or fluid/PEG) and voice development. Results Swallowing was better in the group who had no radiotherapy ( P = 0.0037). There was no difference in voice function between the two groups. We also demonstrated that females had a worse swallowing outcome ( P = 0.0101), as did advanced nodal stage ( P = 0.001). Conclusions RT adversely affects the swallowing results but not the speech results after TL when given either as initial treatment or postoperatively. This should be kept in mind in the decision-making process in the treatment of patients with carcinoma of the larynx.

scholarly journals Convergent and Distinct Effects of Multisensory Combination on Statistical Learning Using a Computer Glove

2021 ◽  
Vol 11 ◽  
Author(s):  
Christopher R. Madan ◽  
Anthony Singhal
Keyword(s):  
Statistical Learning ◽  
Visual Cues ◽  
Explicit Knowledge ◽  
Reaction Time Task ◽  
Musical Instrument ◽  
Implicit Knowledge ◽  
Auditory Cues ◽  
Motor Movements ◽  
Explicit And Implicit Knowledge ◽  
Single Modality

Learning to play a musical instrument involves mapping visual + auditory cues to motor movements and anticipating transitions. Inspired by the serial reaction time task and artificial grammar learning, we investigated explicit and implicit knowledge of statistical learning in a sensorimotor task. Using a between-subjects design with four groups, one group of participants were provided with visual cues and followed along by tapping the corresponding fingertip to their thumb, while using a computer glove. Another group additionally received accompanying auditory tones; the final two groups received sensory (visual or visual + auditory) cues but did not provide a motor response—all together following a 2 × 2 design. Implicit knowledge was measured by response time, whereas explicit knowledge was assessed using probe tests. Findings indicate that explicit knowledge was best with only the single modality, but implicit knowledge was best when all three modalities were involved.

How to create a human communication system

2017 ◽  
Vol 18 (3) ◽  
pp. 314-329 ◽  
Author(s):  
Casey J. Lister ◽  
Nicolas Fay
Keyword(s):  
Theoretical Model ◽  
Communication System ◽  
Communication Systems ◽  
Experimental Studies ◽  
Human Communication ◽  
Content Type ◽  
Sign Systems ◽  
The Creation ◽  
Interactive Feedback ◽  
Single Modality

Following a synthesis of naturalistic and experimental studies of language creation, we propose a theoretical model that describes the process through which human communication systems might arise and evolve. Three key processes are proposed that give rise to effective, efficient and shared human communication systems: (1) motivated signs that directly resemble their meaning facilitate cognitive alignment, improving communication success; (2) behavioral alignment onto an inventory of shared sign-to-meaning mappings bolsters cognitive alignment between interacting partners; (3) sign refinement, through interactive feedback, enhances the efficiency of the evolving communication system. By integrating the findings across a range of diverse studies, we propose a theoretical model of the process through which the earliest human communication systems might have arisen and evolved. Importantly, because our model is not bound to a single modality it can describe the creation of shared sign systems across a range of contexts, informing theories of language creation and evolution.

Sign in / Sign up

Export Citation Format

Share Document