Mudit Garg
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Mudit Garg has a total of 33 co-authors. Here are the top co-authors by number of shared publications:
- Chandan Guha
Department of Radiation Oncology, Albert Einstein College of Medicine, Bronx, NY 10461, USA
4 shared publications - Rafi Kabarriti
Departments of Radiation Oncology, Montefiore Einstein Comprehensive Cancer Center, Bronx, NY
3 shared publications - Bhadrasain Vikram
Radiation Oncology, Albert Einstein College of Medicine, Bronx, NY, USA
2 shared publications P.K. Chakravarty
Radiation Oncology, Albert Einstein College of Medicine, Bronx, NY, USA
2 shared publicationsS. Kalnicki
Department of Radiation Oncology, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY
2 shared publicationsA.M. Asaro
Department of Radiation Oncology, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY
1 shared publication
Latest Publications
- Open Access
- 1 December 2025
- Published by Institute of Advanced Engineering and Science in IAES International Journal of Artificial Intelligence (IJ-AI)
- Vol. 14 (6)
EmoVibe: AI-driven multimodal emotion analysis for mental health via social media dashboards
Monitoring mental health via social media often utilizes unimodal approaches, such as sentiment analysis on text or single-staged image categorization, or executes early feature fusion. However, in real-world contexts where emotions are conveyed via text, emojis, and images, unimodal approach leads to obscured decision-making pathways and overall diminished performance. To overcome these limitations, we propose EmoVibe, a hybrid multimodal AI framework for emotive analysis. EmoVibe uses attention-based late fusion strategy, where text embeddings are generated from bidirectional encoder representations from transformers (BERT) and visual features are extracted by vision transformer. Subsequently, emoticon vectors linked to avatars are processed independently. Later, these independent data features are integrated at higher levels, enhancing interpretability and performance. In contrast to early fusion methods and integrated multimodal large language models (LLMs) like CLIP, Flamingo, GPT-4V, MentaLLaMA, and domain-adapted models like EmoBERTa, EmoVibe preserves modality-specific contexts without premature fusion. This architecture saves processing cost, allowing for clearer, unambiguous rationalization and explanations. EmoVibe outperforms unimodal baselines and early fusion models, obtaining 89.7% accuracy on GoEmotions, FER, and AffectNet, compared to BERT's 87.4% and ResNet-50's 84.2%, respectively. Furthermore, a customizable, real time, privacy-aware dashboard is created, supporting physicians and end users. This technology enables scalable and proactive intervention options and fosters user self-awareness of mental health.
- 1 December 2025
- Published by Institute of Advanced Engineering and Science in IAES International Journal of Artificial Intelligence (IJ-AI)
- Vol. 14 (6)
Open AccessEmoVibe: AI-driven multimodal emotion analysis for mental health via social media dashboards
Monitoring mental health via social media often utilizes unimodal approaches, such as sentiment analysis on text or single-staged image categorization, or executes early feature fusion. However, in real-world contexts where emotions are conveyed via text, emojis, and images, unimodal approach leads to obscured decision-making pathways and overall diminished performance. To overcome these limitations, we propose EmoVibe, a hybrid multimodal AI framework for emotive analysis. EmoVibe uses attention-based late fusion strategy, where text embeddings are generated from bidirectional encoder representations from transformers (BERT) and visual features are extracted by vision transformer. Subsequently, emoticon vectors linked to avatars are processed independently. Later, these independent data features are integrated at higher levels, enhancing interpretability and performance. In contrast to early fusion methods and integrated multimodal large language models (LLMs) like CLIP, Flamingo, GPT-4V, MentaLLaMA, and domain-adapted models like EmoBERTa, EmoVibe preserves modality-specific contexts without premature fusion. This architecture saves processing cost, allowing for clearer, unambiguous rationalization and explanations. EmoVibe outperforms unimodal baselines and early fusion models, obtaining 89.7% accuracy on GoEmotions, FER, and AffectNet, compared to BERT's 87.4% and ResNet-50's 84.2%, respectively. Furthermore, a customizable, real time, privacy-aware dashboard is created, supporting physicians and end users. This technology enables scalable and proactive intervention options and fosters user self-awareness of mental health.
- Open Access
- 1 September 2019
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 105 (1) , E463
Real-time Activity Monitoring to Prevent Admissions during Radiotherapy (RAMPART)
- 1 September 2019
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 105 (1) , E463
Open AccessReal-time Activity Monitoring to Prevent Admissions during Radiotherapy (RAMPART)
- Open Access
- 1 November 2018
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 102 (3) , e172-E173
NLR as a Prognostic Factor in Solid Tumors and in a Transplantable Solid Tumor Mouse Model
- 1 November 2018
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 102 (3) , e172-E173
Open AccessNLR as a Prognostic Factor in Solid Tumors and in a Transplantable Solid Tumor Mouse Model
- Open Access
- 1 November 2015
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 93 (3) , S209-S210
Gene Expression Biomarkers for Radiation Toxicity With a Focus on Mucositis in Head and Neck Cancers
- 1 November 2015
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 93 (3) , S209-S210
Open AccessGene Expression Biomarkers for Radiation Toxicity With a Focus on Mucositis in Head and Neck Cancers
- 1 November 2001
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 51 (3) , 105-106
A novel therapeutic strategy for hepatocellular carcinoma: immunomodulation by Flt3-Ligand (Flt3L) following whole liver irradiation and radio-inducible HSV-TK gene therapy
- 1 November 2001
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 51 (3) , 105-106
A novel therapeutic strategy for hepatocellular carcinoma: immunomodulation by Flt3-Ligand (Flt3L) following whole liver irradiation and radio-inducible HSV-TK gene therapy
- Open Access
- 1 November 2001
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 51 (3) , 181-182
Effect of ionizing radiation (IR) on expression of apoptosis related genes in a rat salivary gland derived cell line: implication of bcl-2 gene therapy in xerostomia
- 1 November 2001
- Published by Elsevier in International Journal of Radiation Oncology - Biology - Physics (IJROBP)
- Vol. 51 (3) , 181-182
Open AccessEffect of ionizing radiation (IR) on expression of apoptosis related genes in a rat salivary gland derived cell line: implication of bcl-2 gene therapy in xerostomia