---
license: mit
datasets:
- sid220/asl-now-fingerspelling
language:
- en
metrics:
- accuracy
library_name: keras
---

# ASLNow!
ASLNow! is a web app designed to make learning ASL fingerspelling easy and fun! You can try it live at [asl-now.vercel.app](https://asl-now.vercel.app/).

Demo: [https://www.youtube.com/watch?v=Wi5tAxVasq8](https://www.youtube.com/watch?v=Wi5tAxVasq8)
## Model
This model, trained on the isolated fingerspelling dataset is licensed under the MIT License. It will be updated frequently as more data is collected.

### Format
![Overview of Model](images/plotted_model.png)

#### Input
21 hand landmarks, each composed of `x`, `y` and `z` coordinates. The `x` and `y` coordinates are normalized
to `[0.0, 1.0]` by the
image width and height, respectively. The `z` coordinate represents the landmark depth, with the depth at the wrist
being
the origin. The smaller the value, the closer the landmark is to the camera. The magnitude of `z` uses roughly the same
scale as x.

![Hand Landmarks](https://developers.google.com/static/mediapipe/images/solutions/hand-landmarks.png)
From: [https://developers.google.com/mediapipe/solutions/vision/hand_landmarker](https://developers.google.com/mediapipe/solutions/vision/hand_landmarker)

Example:

```
[
  # Landmark 1
  [x, y, z],
  # Landmark 2
  [x, y, z],
  ...
  # Landmark 20
  [x, y, z]
  # Landmark 21
  [x, y, z]
]
```

#### Output
The probability of each class, where classes are defined as such:
```json
{
    "A": 0,
    "B": 1,
    "C": 2,
    "D": 3,
    "E": 4,
    "F": 5,
    "G": 6,
    "H": 7,
    "I": 8,
    "J": 9,
    "K": 10,
    "L": 11,
    "M": 12,
    "N": 13,
    "O": 14,
    "P": 15,
    "Q": 16,
    "R": 17,
    "S": 18,
    "T": 19,
    "U": 20,
    "V": 21,
    "W": 22,
    "X": 23,
    "Y": 24,
    "Z": 25
}
```