Pranav Saxena

Hi! I'm Pranav, a senior in ECE at BITS Pilani, Goa, India. I'm interested in 3D Computer Vision, 3D Scene Understanding and Multi-Modal Generative Modelling. Currently, I'm at Carnegie Mellon University's Robotics Institute pursuing my undergraduate thesis under the supervision of Dr. Wenshan Wang at AirLab and Dr. Ji Zhang. Here, I have worked on Vision Language Navigation in Outdoor Scenarios and I am currently working on 3D Object Representation Learning.

Since my sophomore year, I have also been remotely working at MARMoT Lab, NUS under the supervision of Prof. Guillaume Sartoretti. Here, I have worked on 3D Scene Graph Generation using VLMs, Next-Best-View Policy for NeRFs (co-supervised by Prof. Marija Popović), Open-Vocabulary Gaussian Splatting.

At my university, I am grateful to have worked with Prof. Neena Goveas, Prof. Naveen Gupta, Prof. Sarang Dhongdi, and Prof. Tanmay Tulsidas Verlekar.

Beyond academics, I enjoy watching series, movies, travelling and playing sports. I also love tinkering with drones, and I founded an autonomous drone team in my sophomore year ;)

News

October 2025: My Paper, LLM-RG, has been accepted at Human-aware Embodied AI Workshop at IROS 2025!
July 2025: I'm organizing the CMU-VLA Challenge at IROS 2025.
June 2025: My paper, UAV-VLN, has been accepted at European Conference on Mobile Robots 2025 for Oral Presentation!
June 2025: I've joined Carnegie Mellon University as a visiting researcher!
Nov 2024: Won Best Missions Award at an all India autonomous drone competition, SAE Aerothon'24!
April 2024: I've joined NUS as a research intern!
March 2024: I've founded an autonomous drone team, Team Artemis!

Experience

Visiting Researcher

Carnegie Mellon University

Jun 2025 – Present Pittsburgh, Pennsylvania, USA

Pursuing my undergraduate thesis at Carnegie Mellon University’s Robotics Institute under the supervision of Dr. Wenshan Wang and Dr. Ji Zhang
Working on 3D Object Representation Learning
Previously worked on Vision-Language Navigation for Outdoor Scenarios and co-organized the CMU-VLA Challenge

Research Intern

National University of Singapore

Apr 2024 – Present Remote

Research intern at MARMoT Lab, NUS under the supervision of Prof. Guillaume Sartoretti
Worked on 3D Scene Graph Generation using Vision-Language Models
Designed a generalized framework for Open-Vocabulary 3D Gaussian Splatting, eliminating the need for a per-scene autoencoder
Developed a Next-Best-View Policy for 3D Reconstruction using Neural Radiance Fields (co-supervised by Prof. Marija Popović)

Publications

UAV-VLN: End-to-End Vision Language guided Navigation for UAVs
Pranav Saxena, Nishant Raghuvanshi, Neena Goveas
ECMR 2025 (Oral Presentation)
arXiv

UAV-VLN leverages the common-sense reasoning of LLMs and a vision model for cross-modal grounding to plan context-aware aerial trajectories from free-form natural language instructions.

LLM-RG: Referential Grounding in Outdoor Scenarios using Large Language Models
Pranav Saxena, Avigyan Bhattacharya, Ji Zhang, Wenshan Wang
Human-aware Embodied AI Workshop @ IROS 2025
arXiv, Code

A zero-shot hybrid pipeline that leverages Large Language Models for symbolic reasoning and Vision-Language Models for fine-grained attribute extraction to perform robust referential grounding in challenging outdoor driving scenarios.

ZING-3D: Zero-shot Incremental 3D Scene Graphs via Vision-Language Models
Pranav Saxena, Jimmy Chiun
Preprint
arXiv

Zero-shot framework for incremental 3D scene graph generation using vision-language models, enabling open-vocabulary reasoning and geometric grounding for embodied environments.

Gen-LangSplat: Generalized Language Gaussian Splatting with Pre-Trained Feature Compression
Pranav Saxena
Preprint
arXiv, Code

A generalized 3D Gaussian Splatting framework that eliminates per-scene training and improves cross-scene generalization via a ScanNet-trained autoencoder for CLIP feature compression. This work was done in collaboration with NUS and TU Delft.

Other Projects

3D Object Representation Learning
Carnegie Mellon University

Creating a generalized 3D encoder that can be used to represent 3D objects. The learned embeddings could be used for various downstream tasks like 3D Object Segmentation, 3D Shape Completion, 3D Captioning.

Next-Best-View Policy for NeRFs
National University of Singapore, TU Delft, Purdue University
Code

Developed an attention-based Next-Best-View (NBV) policy for Neural Radiance Fields to select the most informative viewpoints for 3D reconstruction. Used a Vision Transformer to encode multi-view image embeddings and candidate poses to predict ΔPSNR gain, supervised with ground-truth from pixelNeRF.

3D Human Pose Estimation using WiFi signals
BITS Pilani, Goa
Report

Developed a two-stage learning framework for 3D human pose estimation using WiFi signals, and incorporated phase sanitization to improve CSI quality. Validated on the MM-Fi dataset, achieving enhanced accuracy and robustness over raw CSI inputs.

Artemis
BITS Pilani, Goa
Video

Developed an autonomous drone using ROS2 and MAVROS as part of my competitive autonomous drone team, Team Artemis. Implemented a robust surveying and informative path planning algorithm. Hardware stack includes a Raspberry Pi 5, PX4 flight controller.

Swarm Bots
BITS Pilani, Goa
Poster

Developed an algorithm for swarm robots by combining Artificial Potential Field and Reciprocal Velocity Obstacles methods to overcome the shortcomings of both methods.

Stock Price Prediction using Optimal Stopping Theory
BITS Pilani, Goa
Code, Report

This is the course project for MATH F424 Applied Stochastic Process. Developed a stock price prediction model using Optimal Stopping Theory and Geometric Brownian Motion to maximize returns by determining the optimal time to sell stocks. Backtested the model on historical NSE stock data over 10 years.