Logo Freelancer
Log Masuk Daftar
Siarkan Projek
MENJELAJAH
C++ Programming
Machine Learning (ML)
Mobile App Development
Software Development
Web Hosting
Gambar muka depan profil
Anda kini mengikuti
Ralat mengikuti pengguna.
Pengguna ini tidak membenarkan pengguna untuk mengikuti mereka.
Anda telah mengikuti pengguna ini.
Pelan keahlian anda hanya membenarkan 0 mengikuti. Naik taraf di sini.
Berjaya menyahikut
Ralat tidak mengikuti pengguna.
Anda berjaya mengesyorkan
Ralat mengesyorkan pengguna.
Sesuatu telah berlaku. Sila segar semula halaman dan cuba lagi.
E-mel berjaya disahkan.
Avatar Pengguna

Pankaj R.

@Pankajrajput0312

0.0
0
0.0
0.0
0%

AI agents Developer | RAG | NLP | computer vision

$10 USD / Jam
・
Bendera
India (10:22 PG)
・
Menyertai pada Mac 22, 2020
$10 USD / Jam
・
Hi, I'm Pankaj, a passionate tech enthusiast with 4 years of experience in Machine Learning, App and web development. As the co-founder and CTO of a shealth startup, I turn vision into reality by building cutting-edge products. To date, I’ve launched 4 live apps, 3 websites, and 2 web apps. In addition to my development work, I have 2.5 years of experience in deep learning, machine learning, and Microsoft Azure. My expertise spans from foundational concepts to advanced topics like Generative Adversarial Networks, Transformers, and Bilinear Graph Neural Networks. I've worked on diverse projects, including self-driving cars, depth estimation, semantic segmentation, and more, all aimed at pushing the boundaries of AI innovation. I'm always excited to connect with like-minded individuals who share a passion for technology and innovation. Let's explore how we can collaborate to create impactful solutions!
Pengesahan
Freelancer Mobile App Developers India
Lihat semua Mobile App Developers dalam India

Portfolio

Portfolio

9328136
Worked in pairs, to predict the relative Depth of objects in the image, using RGB image from Monocular Camera. We build our Custom model and loss for training our model.
Monocular Depth Estimation
Monocular Depth Estimation
12415189
Automated generation of optimized floor plans for residential, commercial, or industrial spaces. Process Input: User specifies dimensions, rooms, and preferences. Model: Utilized GANs and LLMs for layout generation and customization. Output: Delivers 2D and 3D plans with interactive modification features. Features Customizable layouts based on user needs. Optimized for space, lighting, and functionality. Accurate designs aligned with architectural standards.
Floor Pan Desgin Generation using AI
9328141
Objective: Developed a generative model to transform satellite imagery into corresponding street map views using a Pix2Pix GAN architecture. Technical Approach: Employed a UNet model as the Generator to create high-quality street maps from satellite views. Used a PatchGAN model as the Discriminator to assess the realism of the generated outputs. Combined Composite Adversarial Loss and L1 loss for optimization, balancing pixel-level accuracy and visual plausibility. Outcomes: Successfully generated realistic street map views from satellite images, demonstrating high fidelity in spatial and semantic details, making it suitable for urban planning and GIS applications.
Gen Ai Satellite to Street view
9328146
Objective: Developed a system to predict steering angles (in degrees) based on sequential input images captured from a vehicle's front-facing camera to enable autonomous driving. Technical Approach: Designed a deep learning pipeline using a Convolutional Neural Network (CNN) to process image sequences and predict the steering angle. The model was trained with L1 loss to minimize the error between predicted and actual steering angles. Outcomes: Successfully implemented a model capable of real-time steering angle prediction, demonstrating robust performance on test datasets and driving simulation environments.
Self Driving Car
Self Driving Car
9328133
Image Captioning
9328136
Worked in pairs, to predict the relative Depth of objects in the image, using RGB image from Monocular Camera. We build our Custom model and loss for training our model.
Monocular Depth Estimation
Monocular Depth Estimation
12415189
Automated generation of optimized floor plans for residential, commercial, or industrial spaces. Process Input: User specifies dimensions, rooms, and preferences. Model: Utilized GANs and LLMs for layout generation and customization. Output: Delivers 2D and 3D plans with interactive modification features. Features Customizable layouts based on user needs. Optimized for space, lighting, and functionality. Accurate designs aligned with architectural standards.
Floor Pan Desgin Generation using AI
9328141
Objective: Developed a generative model to transform satellite imagery into corresponding street map views using a Pix2Pix GAN architecture. Technical Approach: Employed a UNet model as the Generator to create high-quality street maps from satellite views. Used a PatchGAN model as the Discriminator to assess the realism of the generated outputs. Combined Composite Adversarial Loss and L1 loss for optimization, balancing pixel-level accuracy and visual plausibility. Outcomes: Successfully generated realistic street map views from satellite images, demonstrating high fidelity in spatial and semantic details, making it suitable for urban planning and GIS applications.
Gen Ai Satellite to Street view
9328146
Objective: Developed a system to predict steering angles (in degrees) based on sequential input images captured from a vehicle's front-facing camera to enable autonomous driving. Technical Approach: Designed a deep learning pipeline using a Convolutional Neural Network (CNN) to process image sequences and predict the steering angle. The model was trained with L1 loss to minimize the error between predicted and actual steering angles. Outcomes: Successfully implemented a model capable of real-time steering angle prediction, demonstrating robust performance on test datasets and driving simulation environments.
Self Driving Car
Self Driving Car
9328133
Image Captioning
Perubahan disimpan
0.0 · 0 Reviews
Ulasan
Tiada ulasan untuk dilihat di sini!
Pengalaman
CTO
Jun, 2022 - Sekarang
•
2 tahun, 6 bulan
Echio
Jun, 2022 - Sekarang
•
2 tahun, 6 bulan
● Built the Echio app & Webapp from idea to Development, from building MVP to talking to customers. ● Designed a scalable backend with Node.js (TypeScript) and MongoDB, enabling multi-deliverable campaign management and seamless payment workflows. ● Developed APIs and integrated AI algorithms to streamline influencer discovery, reducing brand collaboration time from 7 days to 24 hours. ● Built a responsive and user-friendly frontend using React.js, and Redux for the State Management.
Bendera
Delhi, India
Jun, 2022 - Sekarang
•
2 tahun, 6 bulan
Machine Learning Intern
Apr, 2022 - Jul, 2022
•
3 bulan, 1 hari
HIKE
Apr, 2022 - Jul, 2022
•
3 bulan, 1 hari
● Implemented Natural Language Understanding based Question Generation from Unstructured text using Transformer Model for Generating Questions and Sense2Vec model for the Distractors Generations using Python ● Created FLASK REST API to generate questions and Dump data into the MongoDB. ● Designed Front-End of the Website using Python Widget and VOILA. ● Generated more than 75K approved questions, which were used as a Database for the Quizzy game.
Bendera
delhi, India
Apr, 2022 - Jul, 2022
•
3 bulan, 1 hari
Machine Learning Internship
Jan, 2021 - Jun, 2021
•
4 bulan, 30 hari
Phidian Tecjnologies Pvt. Ltd
Jan, 2021 - Jun, 2021
•
4 bulan, 30 hari
Work with the Data analyst team on Building Machine learning Models for Companies Sales Predictions. Created multiple Aws lambda functions for Data Visualization in the HR management Platform
Jan, 2021 - Jun, 2021
•
4 bulan, 30 hari
Pendidikan
Delhi College of Engineering
2019 - 2023
•
4 tahun
B.Tech
Bendera
India
2019 - 2023
•
4 tahun
Jemputan berjaya dihantar!
Terima kasih! Kami telah menghantar pautan melalui e-mel kepada anda untuk menuntut kredit percuma anda.
Sesuatu telah berlaku semasa menghantar e-mel anda. Sila cuba lagi.
Logo Freelancer Malaysia / Bahasa Malaysia
Bantuan & Sokongan
 Pekerja bebas
Kategori ProjekPeraduanPekerja BebasEnterprise Innovation Challenges Pembangunan AIKeahlianProgram Pekerja Bebas PilihanPengurusan ProjekPekerjaan TempatanPhoto AnywherePameranAPI untuk PembangunDisahkanAplikasi Desktop
 Tentang
Tentang kamiBagaimana ia BerfungsiKeselamatanPelaburPeta LamanKisahDalam BeritaPasukanAnugerahSiaran AkhbarKerjaya
 Terma
Dasar PrivasiTerma dan SyaratDasar HakciptaTatakelakuanYuran dan Caj
 Rakan Kongsi
Escrow.comLoadshiftWarrior Forum
 Aplikasi
Logo Apple App StoreLogo Mainan Google
Salin ke papan keratan gagal, sila cuba lagi selepas melaraskan kebenaran anda.
Disalin ke papan keratan.
Pengguna Berdaftar Jumlah Pekerjaan Disiarkan
Freelancer ® is a registered Trademark of Freelancer Technology Pty Limited (ACN 142 189 759)
Copyright © 2025 Freelancer Technology Pty Limited (ACN 142 189 759)
Memuatkan pratonton
Kebenaran diberikan untuk Geolocation.
Sesi log masuk anda telah luput dan telah dilog keluar. Sila log masuk sekali lagi.