Gautam Verma

Gautam Verma

Auto-generated profile, based on publicly available information

Gautam Verma

Computer Science and Engineering Department, Amity University, Noida, Uttar Pradesh, India

Research Focus & Metrics

Academic Career Length

11Years (2016-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 March 2025 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
MKMohan Karnati
Mohan Karnati
GSGeet Sahu
Geet Sahu
GVGautam Verma
Gautam Verma
ASAyan Seal
Ayan Seal
MDMalay Kishore Dutta
Malay Kishore Dutta
Joanna Jaworek-KorjakowskaJoanna Jaworek-Korjakowska
Joanna Jaworek-Korjakowska
Despite numerous advances, research on facial expression recognition (FER) has recently been proceeding from confined lab circumstances to in-the-wild environments because it is still a arduous and demanding problem due to occlusion and pose changes, intra-class and intensity variations caused by illumination, and insufficient training data. Moreover, most state-of-the-art approaches typically use the entire face as a homogenous source of features for FER. It is observed from past studies on psychology and physiology that the mouth and eyes reflect the variations of various facial expressions, which are closely related to the manifestation of emotion. A novel method is proposed in this study to address some of the issues mentioned above. Firstly, modified homomorphic filtering is employed to normalize the illumination, which reduces the intra-class and intensity differences. Secondly, the normalized face image is cropped into five local regions to emphasize expression-specific characteristics. Thirdly, a unique blended feature attention network (BFAN) is designed for FER. The first module of the BFAN is devised to improve both high-level (HL) and low-level (LL) spatial structural properties of the individual region and normalized face image, which are fed into a residual dilated multi-scale feature extraction module to retrieve rich contextual information. Furthermore, spatial and channel-wise attention modules are employed to extract the most relevant and discriminative features from the LL and HL features. Then, both feature maps are integrated by performing element-wise addition. These features are inputted into dense layers followed by a softmax layer to compute probability scores. Finally, the Choquet fuzzy integral is applied to the computed probability scores to get the final outcome. The significance and superiority of the proposed method are exemplify by performing several experiments with fourteen existing approaches on six benchmark datasets.

31 July 2016 in Case Reports in Cardiology
Verified
Journal Article
Open access
PDF file
JOJohn P. O’lAughlin
John P. O’lAughlin
GVGautam Verma
Gautam Verma
IGIosif Gulkarov
Iosif Gulkarov
We present a case of a young female with stroke symptoms who underwent valve sparing resection of a presumed fibroelastoma based on echocardiographic findings. After confirming embolic stroke, she underwent excision of the lesion, which on pathology revealed a nonbacterial thrombus. Ultimately, this led to a more extensive work-up leading to the discovery of a papillary serous ovarian carcinoma, the underlying cause of her hypercoagulable state. The initial echocardiographic findings painted the clear picture of a papillary tumor on the aortic valve which was likely the source of the emboli resulting in ischemic stroke. This unique case presentation illustrates that imaging, including echocardiography, may not always coincide with the clinical diagnosis. Thus, understanding the differential diagnoses of cardiac masses is of vital clinical significance. The distinction of fibroelastoma versus the much less common finding of aortic thrombus may lead to early diagnosis of malignancy and prevention of life threatening events due to stroke or undiagnosed disease.