For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
M

Malla Dhan Raj

Automatic Grading and Quality Inspection of Indian Mangoes—Image Classification/AI Labeling

India flagBangalore, India
$5.00/hrEntry LevelOther

Key Skills

Software

Other

Top Subject Matter

Agricultural Image Quality Inspection
Forestry/Environmental Video Analysis
Natural Language Understanding—LLM Evaluation/QA

Top Data Types

ImageImage
VideoVideo
TextText

Top Task Types

ClassificationClassification
SegmentationSegmentation
Fine-tuningFine-tuning

Freelancer Overview

Automatic Grading and Quality Inspection of Indian Mangoes—Image Classification/AI Labeling. Brings 1 year of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other. Education includes Master of Technology and Bachelor of Technology, Indian Institute of Technology Kharagpur (2020) and Higher Secondary Certificate, Narayana Junior College (2020). AI-training focus includes data types such as Image, Video, and Text and labeling workflows including Classification, Segmentation, and Fine-tuning.

Entry LevelEnglishHindiTelugu

Labeling Experience

Video Segmentation/Detection for Tree Trunk Borders—Computer Vision Intern

OtherVideoSegmentation
As a Computer Vision Intern, I developed tree trunk border detection models and segmentation systems for video data. My work centered on creating both orientation and semantic segmentation labels for continuous video frames. I processed and labeled video datasets to train border detection and tracking models. • Used PCA for orientation and segmenting trunk borders in videos. • Applied SAM for semantic segmentation across 300+ videos. • Built optical flow systems for tracking points over time within segmented borders. • Published a PyPI library that includes video segmentation models for trunk detection.

As a Computer Vision Intern, I developed tree trunk border detection models and segmentation systems for video data. My work centered on creating both orientation and semantic segmentation labels for continuous video frames. I processed and labeled video datasets to train border detection and tracking models. • Used PCA for orientation and segmenting trunk borders in videos. • Applied SAM for semantic segmentation across 300+ videos. • Built optical flow systems for tracking points over time within segmented borders. • Published a PyPI library that includes video segmentation models for trunk detection.

2024 - 2024

Image Labeling for Denoising—Devanagari Handwritten Dataset

OtherImageClassification
In a project on image denoising of handwritten Devanagari script images, I generated and labeled datasets by applying various noise types. My responsibilities included curating noisy-clean image pairs to facilitate supervised training for denoising AI models. I managed data preparation workflows that are foundational to image quality restoration AI systems. • Created datasets with Gaussian, Salt and Pepper, Poisson, and Speckle noise. • Generated ground truth and noisy pairs for model training and evaluation. • Enabled effective benchmarking of deep CNN and GAN denoising models. • Contributed labeled image data to advance handwritten script processing tools.

In a project on image denoising of handwritten Devanagari script images, I generated and labeled datasets by applying various noise types. My responsibilities included curating noisy-clean image pairs to facilitate supervised training for denoising AI models. I managed data preparation workflows that are foundational to image quality restoration AI systems. • Created datasets with Gaussian, Salt and Pepper, Poisson, and Speckle noise. • Generated ground truth and noisy pairs for model training and evaluation. • Enabled effective benchmarking of deep CNN and GAN denoising models. • Contributed labeled image data to advance handwritten script processing tools.

2023 - 2023

Text Dataset Annotation and LLM Fine-tuning—Kaggle Competition

OtherTextFine Tuning
During the Kaggle LLM Science Exam competition, I preprocessed datasets and fine-tuned large language models (LLMs) with various NLP techniques. This involved preparing text data for supervised learning and evaluation, as well as prompt design for improved model inference. The experience provided hands-on knowledge of text annotation required for LLM training and evaluation. • Preprocessed and annotated datasets for QA and NER fine-tuning tasks. • Fine-tuned DeBERTa-v2, BERT, and Platypus2 for science exam answers. • Designed prompts and performed evaluation to enhance model MAP@3 score. • Achieved a top ranking by improving annotation and tuning procedures for text models.

During the Kaggle LLM Science Exam competition, I preprocessed datasets and fine-tuned large language models (LLMs) with various NLP techniques. This involved preparing text data for supervised learning and evaluation, as well as prompt design for improved model inference. The experience provided hands-on knowledge of text annotation required for LLM training and evaluation. • Preprocessed and annotated datasets for QA and NER fine-tuning tasks. • Fine-tuned DeBERTa-v2, BERT, and Platypus2 for science exam answers. • Designed prompts and performed evaluation to enhance model MAP@3 score. • Achieved a top ranking by improving annotation and tuning procedures for text models.

2023 - 2023

Automatic Grading and Quality Inspection of Indian Mangoes—Image Classification/AI Labeling

OtherImageClassification
I participated in a machine learning project focused on the automatic grading and quality inspection of Indian mangoes. My role involved extracting foregrounds from images and developing features with computer vision techniques. I then built and trained machine learning models using these features for classification tasks. • Extracted features from mango images using HSV color space and K-means clustering. • Trained Random Forest and SVM models to classify quality grades. • Achieved high accuracy, with F1-scores above 0.93 for Random Forest and 0.95 for SVM. • Contributed to automated visual quality inspection using structured AI labeling workflows.

I participated in a machine learning project focused on the automatic grading and quality inspection of Indian mangoes. My role involved extracting foregrounds from images and developing features with computer vision techniques. I then built and trained machine learning models using these features for classification tasks. • Extracted features from mango images using HSV color space and K-means clustering. • Trained Random Forest and SVM models to classify quality grades. • Achieved high accuracy, with F1-scores above 0.93 for Random Forest and 0.95 for SVM. • Contributed to automated visual quality inspection using structured AI labeling workflows.

2023 - 2023

Education

N

Narayana Junior College

Higher Secondary Certificate, Science

Higher Secondary Certificate
2018 - 2020
B

Bhashyam High School

Secondary School Certificate, General Studies

Secondary School Certificate
2016 - 2018

Work History

I

Indian School of Business

Computer Vision Intern

Hyderabad
2024 - 2024
J

JM Financial

Data Science Summer Intern

Mumbai
2024 - 2024