About
✨ I am currently advancing my academic journey with a master's degree in data science at…
Contributions
-
How can you find the best machine learning algorithms for optimizing robotic decision-making?
First, clarify what your robot needs to do. Understand its tasks and decisions: Does it need quick decisions or can it take its time? Then, consider the data it'll use and how much of it there is. This will help choose algorithms suited for real-time decisions or deeper analysis. Once you know these, pick algorithms that match these needs and test them out. Adjust based on performance until you find the best fit.
Experience
Education
Licenses & Certifications
Volunteer Experience
-
Event Coordinator
VPM's Polytechnic
- 1 month
Arts and Culture
Managed and hosted the Polyspark College-event for 2018 and 2019.
-
Content Management Head
IEEE VIT Student Branch
- 1 year 1 month
Science and Technology
Created engaging visual content and eye-catching infographics for Instagram. Edited and created videos, posters and graphics for events.
Publications
-
RanLayNet: A Dataset for Document Layout Detection used for Domain Adaptation and Generalization
MMAsia '23: Proceedings of the 5th ACM International Conference on Multimedia in Asia
- Spearheaded the creation of RanLayNet, a synthetic document dataset with automated labeling, aimed at enhancing layout detection model adaptability to diverse document formats.
- Conducted empirical experiments demonstrating that a deep layout identification model trained on RanLayNet outperformed models trained solely on real documents, showcasing improved adaptability and robustness.
- Developed domain adaptation approaches to address challenges arising from differences between source…- Spearheaded the creation of RanLayNet, a synthetic document dataset with automated labeling, aimed at enhancing layout detection model adaptability to diverse document formats.
- Conducted empirical experiments demonstrating that a deep layout identification model trained on RanLayNet outperformed models trained solely on real documents, showcasing improved adaptability and robustness.
- Developed domain adaptation approaches to address challenges arising from differences between source and target domains, achieving optimal model performance, as evidenced by a 0.398 and 0.588 mAP95 score in the scientific document domain for the TABLE class.Other authorsSee publication -
GEC-DCL: Grammatical Error Correction Model with Dynamic Context Learning for Paragraphs and Scholarly Papers
International Conference on Big Data and Artificial Intelligence (BDA 2023)
• Developed Dynamic Context Learner (DCL) model GEC-DCL for optimal context breakpoints, outperforming fixed-length models.
• Achieved significant F0.5 score improvements of 77% increase, 19.61% boost, and 10.49% rise, respectively, on CoNLL2014, BEA-Dev, and FCE-Test datasets.
• Validated DCL's performance in scientific writing and scholarly publications with cutting-edge results on S2ORC dataset.Other authorsSee publication -
TC-OCR: TableCraft OCR for Efficient Detection & Recognition of Table Structure & Content
MMIR '23: Proceedings of the 1st International Workshop on Deep Multimodal Learning for Information Retrieval
• Introduced an innovative end-to-end pipeline TC-OCR that integrates state-of-the-art deep learning models, which introduced an application to capture and transfer any table structure and content from one platform to another.
• Achieving a 25% improvement in OCR accuracy, with an impressive IOU of 0.96 for Table class.Other authorsSee publication -
Exploring Controllable Text Generation for Scientific Literature
Arxiv
Our Contributions
- Introduces a novel schema for organizing previous work on controllable text generation.
- Identifies seven crucial components in the text creation process.
- Focuses on controlled generation for scientific literature.
- Describes modulation strategies for each of the seven components.
- Presents a theoretical analysis and qualitative evaluation of these modulation methods.
- Provides insights into potential new architectures by combining these…Our Contributions
- Introduces a novel schema for organizing previous work on controllable text generation.
- Identifies seven crucial components in the text creation process.
- Focuses on controlled generation for scientific literature.
- Describes modulation strategies for each of the seven components.
- Presents a theoretical analysis and qualitative evaluation of these modulation methods.
- Provides insights into potential new architectures by combining these components.
- Anticipates future research that will empirically compare the strengths and utility of these methods.Other authorsSee publication -
Certified Zeroth-order Black-Box Defense with Robust UNet Denoiser
Arxiv
Our contributions:
- Propose a novel black-box denoised smoothing (DS) defense mechanism, ZO-RUDS, by prepending our RDUNet to the black-box model, ensuring black-box defense.
- Also, propose ZO-AE-RUDS in which RDUNet followed by autoencoder (AE) is prepended to the black-box model.Other authorsSee publication
Courses
-
Audit Course on Industry ready Web developer
-
-
Spatial and Geospatial Data Science
-
Projects
-
Video Inpainting
As in my Master's Thesis, I am exploring advanced techniques and algorithms to seamlessly reconstruct missing or damaged portions of videos. Through rigorous research and experimentation, I aim to contribute to the advancement of video inpainting technology, unlocking new possibilities in video editing, restoration, and beyond.
-
Sign Detection System Using Python OpenCV
-
Sign Detection System created using Python and Open CV
An Image processing system that analyzes the hand gestures and provides respective output as per user requirements and it can help to discard the communication barriers
Key Features :
1. Image Detection
2. Image Analysis
3. Algorithm
4. Data Conversion
5. Flexible to useOther creatorsSee project
Honors & Awards
-
BEST PAPER in ICRAET-2023
Institute For Engineering Research and Publication (IFERP)
Test Scores
-
IELTS
Score: 7
Languages
-
English
Full professional proficiency
-
Hindi
Native or bilingual proficiency
-
Marathi
Native or bilingual proficiency
Organizations
-
IEEE
Member
-
People also viewed
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Siddhesh Bangar
-
Siddhesh Bangar
Data Analyst Intern at SpareProvider.com | Passionate Data Science Learner | Front End Developer | Exploring the Intersection of Technology and Finance | Finance Aficionado
-
Siddhesh Bangar
Attended University of Mumbai. Pursuing bachelor of commerce |
-
Siddhesh Bangar
ACCA Affiliate, B.com, M.com
-
Siddhesh Bangar
Student at University of Mumbai
6 others named Siddhesh Bangar are on LinkedIn
See others named Siddhesh Bangar