Jhony Kaesemodel Pontes

AI experiments

Before Music • see music as pressure before the mind turns it into experience
The Space Between • encounter the double-slit experiment and four realities built around it
The Seeing Machine • find the holes, edits, and useful inventions inside visual experience
Inside One Cell • open one gene and follow information until it becomes structure
Two, Apart • run a Bell experiment and test the limits of local explanation
Clockwork Sky • turn one 3D model through days, seasons, phases, eclipses, and tides
When Giants Walked • cross dinosaur worlds separated by impossible stretches of time
Two Epics • untangle the Iliad and Odyssey by story order, memory, and consequence
Gravity Garden • throw a world sideways and learn why orbit is continuous falling
How a Thought Moves • turn many tiny neural inputs into one traveling pulse
The Breath Between Worlds • follow one carbon atom across life, ocean, and deep time
A Clock That Disagrees • make time dilation emerge from a pulse of light
The Probability Machine • watch crowds become stable while the next event stays unknown
How a Machine Sees • transform sensor evidence into hypotheses, motion, and uncertainty
Before Words • trace one word backward through older voices to the edge of known ancestry
The Unwritten • walk the tiny written edge of human time and meet the evidence that survived before it

Publications

What Matters in Range View 3D Object Detection • CoRL 2024 • pdf
Fast Neural Scene Flow • ICCV 2023 • pdf • code • video
Neural Prior for Trajectory Estimation • CVPR 2022 • pdf • supp
Neural Scene Flow Prior • NeurIPS 2021 (Spotlight) • pdf • code • video
Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting • NeurIPS 2021 • pdf • code • project
PointNetLK Revisited • CVPR 2021 (Oral) • pdf • code • video
Scene Flow from Point Clouds with or without Learning • 3DV 2020 (Oral) • pdf
Representation and Reconstruction of 3D Shapes in Computer Vision - PhD Thesis 2020 • pdf
Implicit Surface Representations as Layers in Neural Networks • ICCV 2019 • pdf • code
Image2Mesh: A Learning Framework for Single Image 3D Reconstruction • ACCV 2018 • pdf • code
Learning Free-Form Deformations for 3D Object Reconstruction • ACCV 2018 • pdf • code
Compact Model Representation for 3D Reconstruction • 3DV 2017 • pdf • code
Two-Stage Facial Age Prediction Using Group-Specific Features • ICASSP 2017 • pdf
A Flexible Hierarchical Approach for Facial Age Estimation Based on Multiple Features • Pattern Recognition 2016 • pdf