Skip to content
ModelListedExternal
Download

Hand Gesture Yolo26n

Fine-tuned YOLO26n for 12-class hand gesture recognition, trained on the HaGRID dataset.

Sourcehuggingface.co/SuperSura/hand-gesture-yolo26n/tree/26cab531b83cdc9f09808a73c0dc485e35920c6b
KindOtherVersionRevision 26cab53

Model source

Kind
Other
File size
5.14 MB
Version
vRevision 26cab53
Updated
Sep 16, 2026
Downloads
165
Tasks
Object detection
Source
Hugging Face

Source description

Hand Gesture Recognition — YOLO26n

Fine-tuned YOLO26n for 12-class hand gesture recognition, trained on the HaGRID dataset.

Model Details

AttributeValue
ArchitectureYOLO26n (Ultralytics)
Classes12
Input Size800×800
Batch48
Epochs100
mAP500.994
mAP50-950.862
Weight Size5.1 MB (FP32)

Classes

0: like      1: palm      2: fist       3: peace
4: ok        5: one       6: three      7: four
8: call      9: point    10: grip      11: middle_finger

Usage

from ultralytics import YOLO

model = YOLO("yolo26n_hand-gesture_12c_v1.pt")
results = model("hand.jpg")
results[0].show()

Training Parameters

  • imgsz: 800
  • batch: 48
  • optimizer: auto (AdamW)
  • device: CUDA (NVIDIA RTX 5070 Ti, 16 GB)
  • cache: disk
  • workers: 12
  • patience: 0

Dataset

Trained on a subset of HaGRID (Hand Gesture Recognition Image Dataset), containing ~36,000 images across 12 gesture classes.

Performance Notes

The model converged around epoch 84 (mAP50-95=0.862) and remained stable through epoch 100. Suitable for real-time inference on consumer GPUs.

Sources

1 source
Hugging FacePrimaryVersion 26cab531b83cdc9f09808a73c0dc485e35920c6b
huggingface.co/SuperSura/hand-gesture-yolo26n/tree/26cab531b83cdc9f09808a73c0dc485e35920c6b

Verified Sep 16

Model artifacts

1 artifact

yolo26n_hand-gesture_12c_v1.pt

pt · 5.14 MB · SHA-256 5695ba86a2a6…abc9 · Hugging Face

Download

Source excerpts

2 excerpts
ReadmeSep 16huggingface.co/SuperSura/hand-gesture-yolo26n

--- language: en license: apache-2.0 tags: - hand-gesture - yolo26n - ultralytics - object-detection datasets: - hagrid library_name: ultralytics --- # Hand Gesture Recognition — YOLO26n Fine-tuned YOLO26n for 12-class hand gesture recognition, trained on the HaGRID dataset. ## Model Details | Attribute | Value | |-----------|-------| | **Architecture** | YOLO26n (Ultralytics) | | **Classes** | 12 | | **Input Size** | 800×800 | | **Batch** | 48 | | **Epochs** | 100 | | **mAP50** | 0.994 | | **mAP50-95** | 0.862 | | **Weight Size** | 5.1 MB (FP32) | ## Classes ``` 0: like 1: palm 2: fist 3: peace 4: ok 5: one 6: three 7: four 8: call 9: point 10: grip 11: middle_finger ``` ## Usage ```python from ultralytics import YOLO model = YOLO("yolo26n_hand-gesture_12c_v1.pt") results = model("hand.jpg") results[0].show() ``` ## Training Parameters - `imgsz`: 800 - `batch`: 48 - `optimizer`: auto (AdamW) - `device`: CUDA (NVIDIA RTX 5070 Ti, 16 GB) - `cache`: disk - `workers`: 12 - `patience`: 0 ## Dataset Trained on a subset of [HaGRID](https://hagrid.tilda.ws/) (Hand Gesture Recognition Image Dataset), containing ~36,000 images across 12 gesture classes. ## Performance Notes The model converged around epoch 84 (mAP50-95=0.862) and remained stable through epoch 100. Suitable for real-time inference on consumer GPUs.

NoteSep 16huggingface.co/SuperSura/hand-gesture-yolo26n

Source context: 165 downloads · 0 likes · Pipeline object-detection · Library ultralytics · Repo SuperSura/hand-gesture-yolo26n