Mirror SAM2 ONNX exports - Stage 6
Browse files- README.md +40 -0
- hiera-base-plus/decoder.onnx +3 -0
- hiera-base-plus/encoder.onnx +3 -0
- hiera-tiny/decoder.onnx +3 -0
- hiera-tiny/encoder.onnx +3 -0
README.md
ADDED
|
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- sam2
|
| 5 |
+
- segment-anything
|
| 6 |
+
- onnx
|
| 7 |
+
- mirror
|
| 8 |
+
library_name: onnxruntime
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# SAM2 ONNX Mirror (A.I.M.I)
|
| 12 |
+
|
| 13 |
+
Mirror of [SharpAI's SAM2 ONNX exports](https://huggingface.co/SharpAI) for A.I.M.I's "Smart-Mask" (click-to-select) feature. Used by the Edit tab's SAM2 workflow — user clicks anywhere on a subject, gets a pixel-perfect alpha mask in ~500 ms.
|
| 14 |
+
|
| 15 |
+
Contents unmodified from upstream.
|
| 16 |
+
|
| 17 |
+
## Files
|
| 18 |
+
|
| 19 |
+
### `hiera-tiny/` — small + fast (default at all tiers)
|
| 20 |
+
| File | Size | Purpose |
|
| 21 |
+
|---|---|---|
|
| 22 |
+
| `encoder.onnx` | 128 MB | Image encoder (run once per image) |
|
| 23 |
+
| `decoder.onnx` | 20 MB | Click-to-mask decoder (run per click) |
|
| 24 |
+
|
| 25 |
+
### `hiera-base-plus/` — slightly better masks, slightly slower (Expert option)
|
| 26 |
+
| File | Size | Purpose |
|
| 27 |
+
|---|---|---|
|
| 28 |
+
| `encoder.onnx` | 324 MB | Image encoder (base-plus variant) |
|
| 29 |
+
| `decoder.onnx` | 20 MB | Click-to-mask decoder |
|
| 30 |
+
|
| 31 |
+
Total: ~500 MB.
|
| 32 |
+
|
| 33 |
+
## License
|
| 34 |
+
|
| 35 |
+
Meta AI's Segment Anything 2 (SAM2) code + weights: **Apache 2.0**. ONNX conversion by [SharpAI](https://huggingface.co/SharpAI) on HuggingFace, redistributed here under the same license.
|
| 36 |
+
|
| 37 |
+
## Attribution
|
| 38 |
+
|
| 39 |
+
- **SAM2**: Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman Rädle, Chloe Rolland, Laura Gustafson, Eric Mintun, Junting Pan, Kalyan Vasudev Alwala, Nicolas Carion, Chao-Yuan Wu, Ross Girshick, Piotr Dollár, Christoph Feichtenhofer — *SAM 2: Segment Anything in Images and Videos* (Meta AI Research, 2024).
|
| 40 |
+
- **ONNX export**: SharpAI on HuggingFace.
|
hiera-base-plus/decoder.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90dbe3d746cfdb9050695a94cd3cdd90244c7af48383df2ec340f80587b61726
|
| 3 |
+
size 20639854
|
hiera-base-plus/encoder.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:063717f7b501e7433f775408eafbfcee5684dd19f7f7604fcb81285123db629f
|
| 3 |
+
size 339780953
|
hiera-tiny/decoder.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63198f1f1e273d8f2f4a9d1baf926e53a01d78dc50e0674640e1513dc00d9927
|
| 3 |
+
size 20639854
|
hiera-tiny/encoder.onnx
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df265cb552475e1b3a6cb57c939e57c95ed849bfc2f985c06efab85d8bca6db9
|
| 3 |
+
size 134261315
|