Spaces:
No application file
No application file
Update README.md
Browse fileschanges to org card
README.md
CHANGED
|
@@ -10,7 +10,7 @@ pinned: false
|
|
| 10 |
---
|
| 11 |
|
| 12 |
# Josh Talks AI ποΈ
|
| 13 |
-
*
|
| 14 |
|
| 15 |
Welcome to the official Hugging Face organization for **Josh Talks AI** β a data-centric initiative focused on creating high-quality, culturally-grounded datasets from and for India.
|
| 16 |
|
|
@@ -18,10 +18,23 @@ Welcome to the official Hugging Face organization for **Josh Talks AI** β a da
|
|
| 18 |
|
| 19 |
---
|
| 20 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
## π§ Our Mission
|
| 22 |
|
| 23 |
We are building open datasets to power the next generation of AI tools for India β across languages, domains, and communities. From speech and text to video and behavioral data, we aim to represent the true diversity of the Indian experience.
|
| 24 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
## π¦ Coming Soon
|
| 26 |
|
| 27 |
We are currently preparing a series of open datasets, including:
|
|
@@ -39,7 +52,7 @@ Stay tuned β our first releases are just around the corner!
|
|
| 39 |
|
| 40 |
## π€ Let's Collaborate
|
| 41 |
|
| 42 |
-
We welcome partnerships with researchers, universities, NGOs, and AI developers working on
|
| 43 |
|
| 44 |
---
|
| 45 |
|
|
|
|
| 10 |
---
|
| 11 |
|
| 12 |
# Josh Talks AI ποΈ
|
| 13 |
+
*On a mission to make machines talk like humans.*
|
| 14 |
|
| 15 |
Welcome to the official Hugging Face organization for **Josh Talks AI** β a data-centric initiative focused on creating high-quality, culturally-grounded datasets from and for India.
|
| 16 |
|
|
|
|
| 18 |
|
| 19 |
---
|
| 20 |
|
| 21 |
+
## π§ Problems with ASR
|
| 22 |
+
*Conversational multi speaker speech*
|
| 23 |
+
*Dialect and Accented speech*
|
| 24 |
+
*Child Speech*
|
| 25 |
+
*Small models for real time low latency*
|
| 26 |
+
|
| 27 |
## π§ Our Mission
|
| 28 |
|
| 29 |
We are building open datasets to power the next generation of AI tools for India β across languages, domains, and communities. From speech and text to video and behavioral data, we aim to represent the true diversity of the Indian experience.
|
| 30 |
|
| 31 |
+
## π§ Our Work
|
| 32 |
+
|
| 33 |
+
**Datasets** - Highest quality scientifically designed datasets for training models.
|
| 34 |
+
|
| 35 |
+
**Benchmarks** - We can only fix what we can measure. Our benchmarks tell researchers exactly where their models fail.
|
| 36 |
+
|
| 37 |
+
|
| 38 |
## π¦ Coming Soon
|
| 39 |
|
| 40 |
We are currently preparing a series of open datasets, including:
|
|
|
|
| 52 |
|
| 53 |
## π€ Let's Collaborate
|
| 54 |
|
| 55 |
+
We welcome partnerships with researchers, universities, NGOs, and AI developers working on ASR challenges. If you're interested in contributing or using our datasets, reach out below.
|
| 56 |
|
| 57 |
---
|
| 58 |
|