Available to hire
I am Himanshu Kumar Parashar, a research-oriented computer science student at CURAJ, Ajmer, with hands-on experience in deep learning, medical image analysis, histopathological image segmentation, biomedical signal processing, and time-series modeling. I have contributed to multiple research internships across premier Indian institutes and am developing end-to-end deep learning pipelines for feature extraction, segmentation, and classification. I am pursuing an integrated MSc in Computer Science with a focus on liver disease classification using histopathological image analysis. I am seeking PhD or research positions abroad to advance AI-driven healthcare, computational pathology, and computer vision.
Skills
Language
English
Advanced
Hindi
Advanced
Work Experience
Deep Learning-Based Human Activity Recognition at IIIT Kurnool
May 1, 2025 - July 31, 2025Developed deep learning models including CNN, LSTM, GRU, BiLSTM, Transformer, and RBFNN to perform 6-class human activity recognition using smartphone sensor data. Processed raw accelerometer and gyroscope time-series data for model training and evaluation. Achieved highest performance with a BiLSTM-Transformer hybrid model with training accuracy of 97.85%, validation accuracy of 96.40%, and testing accuracy of 95.80%. Evaluated models using classification reports, confusion matrices, and learning curves to ensure reliable interpretation.
Language Model Trainer at Outliar AI
January 1, 2024 - December 31, 2025Worked on development and fine-tuning of large language models (LLMs) for multiple language domains. Handled localization ranking, prompt generation, and multilingual preference rating for LLMs. Maintained accuracy of over 90% while completing tasks and generating reports within tight deadlines.
Freelancer AI Trainer at Appen Remote
January 1, 2024 - June 9, 2024Contributed to AI training tasks including data annotation and language model optimization. Collaborated on multilingual AI projects to ensure language models accurately reflected diverse linguistic nuances. Conducted quality assurance by reviewing AI model outputs and providing feedback to improve accuracy and relevancy.
Research Intern – Oral Cancer Histopathological Image Analysis at National Institute of Technology (NIT) Jamshedpur
December 1, 2025 - PresentDeveloping deep learning models for automated segmentation and classification of oral squamous cell carcinoma (OSCC) from histopathological whole-slide images. Designing feature extraction pipelines using pre-trained CNNs (ResNet, EfficientNet, DenseNet) and vision transformers. Implementing semantic segmentation architectures (U-Net, Attention U‑Net, DeepLabV3+) for delineation of tumour, stroma, and epithelial regions. Applying stain normalization (Macenko, Vahadane), patch tiling, and augmentation to handle multi-centre dataset variability. Evaluating segmentation with Dice, IoU, and pixel accuracy; performing ablations across backbones.
Research Intern – EEG Signal Analysis for Autism Detection at Indian Institute of Technology (IIT) Kharagpur
May 1, 2025 - November 1, 2025Investigated EEG-based biomarkers for autism spectrum disorder (ASD) detection using the BCIAUT-P300 dataset in a P300 BCI paradigm. Applied signal preprocessing (band-pass filtering, EOG/EMG artifact removal), epoch segmentation, and feature extraction to improve SNR. Designed and trained deep learning classifiers for P300 ERP detection, achieving robust discrimination between target and non-target stimuli. Validated performance with accuracy metrics, confusion matrices, precision-recall analysis, and statistical significance testing.
Deep Learning Research Intern – Human Activity Recognition at Indian Institute of Information Technology, Design and Manufacturing (IIITDM), Kurnool
May 1, 2025 - July 1, 2025Designed and benchmarked six deep learning architectures (CNN, LSTM, GRU, BiLSTM, Transformer, RBFNN) for 6-class human activity recognition from smartphone inertial sensor data. Engineered preprocessing with noise filtering, z-score normalization, and sliding window segmentation. Achieved state-of-the-art performance with a BiLSTM–Transformer hybrid model (Test accuracy: 95.80%), outperforming all baselines. Conducted systematic model comparison using classification reports, confusion matrices, and learning curve analysis.
Language Model Trainer at Outlier AI (Remote)
January 1, 2024 - PresentContributed to large language model fine-tuning through prompt engineering, multilingual evaluation, and preference ranking. Maintained 90%+ task accuracy across evaluation tasks and collaborated on data curation and experimental logging.
AI Data Annotator at Appen (Remote)
January 1, 2024 - PresentPerformed data annotation, quality assurance, and multilingual evaluation for AI training pipelines, supporting tasks across diverse datasets and domains.
Education
Integrated M.Sc Computer Science at Central University of Rajasthan, Ajmer
January 1, 2021 - December 31, 2026Class 12 at Jawahar Navodaya Vidyalaya, Muzaffarpur Bihar
January 1, 2020 - December 31, 2020Matriculation at Ambika Bhawani Public School, Muzaffarpur Bihar
January 1, 2018 - December 31, 2018Integrated MSc in Computer Science at Central University of Rajasthan (CURAJ), Ajmer, India
January 1, 2021 - January 1, 2026Higher Secondary (CBSE) at Jawahar Navodaya Vidyalaya, Muzaffarpur, Bihar, India
January 1, 2020 - March 6, 2026Qualifications
Python Programming : Udemy Certification
January 1, 2022 - December 31, 2022Machine Learning A-Z: Coursera Certification
January 1, 2023 - December 31, 2023Industry Experience
Software & Internet, Healthcare, Life Sciences, Professional Services, Education
Skills
Hire a Online Trainer
We have the best online trainer experts on Twine. Hire a online trainer in Ajmer today.