AI VIETNAM ✦ AIO 2026 cohort ✦ Anno MMXXVI

From the classroom to CVPR, MICCAI & the clinic.

AIMA Research is where Vietnamese students become researchers in translational AI for health. In two years our groups have published 47 papers. This page is the record, and the invitation to AIO 2026.

I

The Record

Two years, forty-seven papers.

Every number below is counted from the paper list on this page. Tap a bar to open that year in the library.

0papers since 2024
0journal articles
0orals & highlights
0distinct venues
33institutions in our network

Papers per year

Conference & workshop Journal
View as table
YearConferenceJournalTotal

Journal impact factors

One bar per journal article. Tap to read.

Where we publish

Sized by paper count. Tap a venue to filter the library.

Conferences & workshops

Journals

II

Cabinet of 2026

Moments worth framing.

Drag or scroll sideways through the year's highlights.

No. 01

CVPR 2026 · Highlight

Unsupervised cryo-ET segmentation with a Stable Diffusion foundation model

Multi-scale segmentation of cellular cryo-electron tomograms, with no labels.

Read paper →

No. 02

ISBI 2026 · CXR-LT Challenge

Winner in both tracks, Long-Tailed and Zero-Shot

Chest X-ray classification when most diseases have few or no training examples.

No. 03

ISBI 2026 · 5 orals

Five papers presented as oral talks

Scribble supervision, SAM on unlabeled images, histopathology prototypes, white blood cells and chest X-rays.

See the orals →

No. 04

npj Digital Medicine · IF 12.4

Robust prediction of drug interactions from chemical descriptors

Our highest-impact journal paper to date.

Read paper →

No. 05

Medical Image Analysis · IF 11.8

Who Gets Missed in the Tail?

Which patient subgroups a long-tailed chest X-ray classifier underdiagnoses.

Read paper →

No. 06

ECCV 2026

ESC: Emotional Self-Correction for reliable vision-language models

Vision-language models that catch and correct their own mistakes.

Read paper →

No. 07

Pattern Recognition · IF 7.6

Switching Dual-Student framework

Adaptive knowledge transfer for semi-supervised medical image segmentation.

Read paper →
III

The Library

Every paper, on one shelf.

Search, filter and cite. Names in gold are AIMA mentors.

    IV

    Atlas of Themes

    What our papers are about.

    A paper can sit in more than one theme. Hover a theme to trace its papers; tap it to open them in the library.

    V

    For AIO 2026

    The path from student to author.

    The same route the authors above took. Most of them were AI VIETNAM students first.

    1. i

      Apply

      When the cohort opens, pick a project prototype and tell us your background and what you want to learn.

    2. ii

      Join a group

      A research group of 4–6 members, led by a PhD student, postdoc or professor in the US or Australia.

    3. iii

      Train

      Read papers critically, build reproducible PyTorch pipelines, handle medical images, and use AI coding agents without losing rigor.

    4. iv

      Do the research

      Own a question, run the experiments, defend the results in weekly meetings. Mentors guide; they do not do it for you.

    5. v

      Publish

      Workshop, then conference, then journal. A record for graduate school and a network that lasts.

    The last cohort: AIO 2025

    Six projects, now closed. They show the scale and targets to expect.

    VI

    The Mentors

    Led by people who have done it before.

    Continue the story

    Themes, projects, people
    and the map of our network.

    Enter AIMA Research