Cedric Kiplimo

About Me

I am a Junior Research Fellow at the Centre for Data Science and Artificial Intelligence (DSAIL), Dedan Kimathi University of Technology, where I work under the guidance of Prof. Ciira wa Maina. My research focuses on building AI methods for application in on two main areas: environmental conservation and agriculture.

In environmental conservation, I develop machine learning methods for scalable forest monitoring. This includes using computer vision to automate field-based forest data collection and employing remote sensing to evaluate the growth of reforested areas.

In agriculture, I collaborate with colleagues to analyze time-series satellite imagery, studying how changes in vegetation indices over time can be used to map individual crop fields, identify crops, and estimate yields. This research addresses a critical need in Kenya, where comprehensive data on crop fields and yields is limited.

Previously, I was a research intern at DSAIL, where I worked on TreeVision, a software tool designed to estimate the biophysical properties of trees using stereoscopic vision. This work formed the basis of my MSc thesis in Electrical Engineering at Dedan Kimathi University of Technology, supervised by Prof. Ciira.

My primary research interests include computer vision, particularly 3D perception of the environment, and probabilistic machine learning.

In my free time, I enjoy learning in a more relaxed way, such as reading or watching documentaries. Currently, Im reading Slow Productivity: The Lost Art of Accomplishment Without Burnout by Cal Newport. I'm also an avid hiker—feel free to ask me about my favorite trails!

Research Interests

Computer Vision: I'm especially interested in 3D perception of the environment, photogrametry, etc. Generally interested in image recognition and image generation
Probablistic machine Learning: Decision theory, neural networks, recommender systems

News

[Feb. 2024] Our Treevision paper gets accepted to Elsevier SoftwareX.
[Mar. 2024] Our paper on compter vision-based low-cost non-contact forest inventory gets accepted to MDPI Challenges.
[Sep. 2024] I attended IEEE ICRA@40 celebrations at Wits, Johannesburg.
[Oct. 2024] Our abstract on forest monitoring using deep learning gets accepted to AGU 2024!
[Dec. 2024] I attended AGU 2024 in Washington DC.

Publications

DSAIL-TreeVision: A software tool for extracting tree biophysical parameters from stereoscopic images

Cedric Kiplimo*, Collins Emasi Epege, Ciira wa Maina, Billy Okal (*Corresponding authors)

SoftwareX, Elsevier BV.

PDF Code Article

Low-Cost Non-Contact Forest Inventory: A Case Study of Kieni Forest in Kenya

Cedric Kiplimo*, Ciira wa Maina, Billy Okal (*Corresponding authors)

Challenges, MDPI.

PDF Code Article
Automating Tree Inventory Using Stereoscopic Vision

Cedric Kiplimo, Ciira wa Maina, Billy Okal

ICRA@40 Africa, IEEE.

Poster presentation
Using Drone Imagery and Deep Learning to Monitor a Reforested Stand in Kenya

Cedric Kiplimo, Samuel Gachana, Viola Orina, Ciira wa Maina, Arthur Sichangi

AGU 2024, AGU.

Abstract, Poster presentation

Guest Lectures

[Nov. 2024] Spatial filtering. ETI 5207 - Digital Image Processing, DeKUT.
[Nov. 2024] Frequency domain filtering. ETI 5207 - Digital Image Processing, DeKUT.
[Dec. 2024] Colour image processing. ETI 5207 - Digital Image Processing, DeKUT.
[Dec. 2024] Image classification. ETI 5207 - Digital Image Processing, DeKUT.

Features

The work I do with the TreeVision software in enhancing forest conservation in Kenya by automating tree inventory was featured on DW TV. Watch the feature here.
Recent feature by Voice of America. Watch here.