Final Project — Multimodal Classification Challenge

CS3317: Artificial Intelligence Instructor: Dr. Tao Huang Metric: F1-macro 2000 test examples · 18 classes
Teams
17
16 with a valid submission
Submissions
443
of 850 possible
Top F1-macro
0.6955
by Proteus

Live leaderboard

Ranked by best F1-macro across all valid submissions per team. Auto-refreshes every 30 seconds. Ties broken by who reached the score first.
Closed The leaderboard closed on 2026-06-12 20:00 CST; no further submissions accepted. Final scores below are final.

Rank Team F1-macro F1-micro Submissions Best at (Beijing)
1
P
Proteus
0.6955 0.7280 36 2026-06-12 16:57
2
L
lastrivia
0.6642 0.7036 28 2026-06-12 19:44
3
AG
Abelian Group
0.6638 0.7029 49 2026-06-12 19:57
4
U
Unmask
0.6422 0.6830 34 2026-06-12 14:07
5
V
vme50
0.6418 0.6844 27 2026-06-12 02:24
6
J
Jubamber
0.6385 0.6857 35 2026-06-09 20:32
7
C
Cortex
0.6384 0.6789 50 2026-06-12 18:02
8
H
Hajimi
0.6348 0.6793 42 2026-06-12 19:59
9
G
ggg
0.6302 0.6737 24 2026-06-12 19:40
10
Q
qwerty
0.6276 0.6678 32 2026-06-12 19:49
11
H
hmmmmmm
0.6195 0.6631 7 2026-06-05 15:43
12
K
KSKBL
0.6193 0.6644 19 2026-05-28 16:52
13
L
lll
0.6151 0.6573 31 2026-06-11 21:26
14
LD
Last Dance
0.6120 0.6646 5 2026-06-05 16:20
15
D
Doctor
0.6008 0.6465 13 2026-06-12 15:35
16
K
kksk
0.5524 0.6102 11 2026-06-11 23:51
17
J
jibao
0

How it works

  1. Register with a group nickname, password, and 1–3 student ID/name pairs.
  2. Log in. The team dashboard shows your remaining submissions for today and overall.
  3. Upload a CSV with one row per test example: example_id,label_0,...,label_17 (values in [0, 1] for probabilities, threshold defaults to 0.5).
  4. Each upload is scored against the held-out labels; F1-macro is the leaderboard metric.
  5. Best score per team determines the rank.