Object Detection
English
document-ai
form-field-detection
hung-k-nguyen commited on
Commit
cee6d22
Β·
0 Parent(s):

form-field-v1 release

Browse files
Files changed (2) hide show
  1. .gitattributes +35 -0
  2. README.md +77 -0
.gitattributes ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: nutrient-commercial
4
+ pipeline_tag: object-detection
5
+ language: [en]
6
+ tags:
7
+ - object-detection
8
+ - document-ai
9
+ - form-field-detection
10
+ datasets:
11
+ - nutrientdocs/form-field-v1-benchmark
12
+ metrics:
13
+ - accuracy
14
+ ---
15
+
16
+ # form-field-v1-medium Β· _commercial_
17
+
18
+ **The highest-accuracy tier of the `form-field-v1` detector family** β€” locates `Text`, `Choice`
19
+ (checkbox/radio), and `Signature` widgets on **empty, filled, and handwritten** document pages. ~34M params,
20
+ NMS-free, single-resolution 1216.
21
+
22
+ The runnable weights are commercial and are not downloadable from this repository β€” this public page is the
23
+ product specification and scorecard. Prefer a small, open model? See
24
+ [`form-field-v1-nano`](https://huggingface.co/nutrientdocs/form-field-v1-nano) (Apache-2.0).
25
+
26
+ - 🎯 **Try it:** [form-field-v1-demo](https://huggingface.co/spaces/nutrientdocs/form-field-v1-demo)
27
+ - πŸ† **Leaderboard:** [form-field-v1-leaderboard](https://huggingface.co/spaces/nutrientdocs/form-field-v1-leaderboard)
28
+ - πŸ“Š **Benchmark:** [form-field-v1-benchmark](https://huggingface.co/datasets/nutrientdocs/form-field-v1-benchmark)
29
+
30
+ ## Results
31
+
32
+ COCO **mAP50-95** on the [`form-field-v1-benchmark`](https://huggingface.co/datasets/nutrientdocs/form-field-v1-benchmark),
33
+ one shared pycocotools scorer across all models.
34
+
35
+ | Model | Empty | Filled | Handwritten |
36
+ |---|---|---|---|
37
+ | **form-field-v1-medium** (this model) | **0.526** | **0.776** | **0.726** |
38
+ | FFDetr (baseline) | 0.461 | 0.170 | 0.225 |
39
+ | FFDNet-L (baseline) | 0.373 | 0.285 | 0.285 |
40
+
41
+ Per-class on empty (mAP50-95): Text 0.612 / Choice 0.601 / Signature 0.364. Full per-class breakdowns are on the
42
+ [leaderboard](https://huggingface.co/spaces/nutrientdocs/form-field-v1-leaderboard).
43
+
44
+ ## Output
45
+
46
+ ```json
47
+ [ { "box": [82, 164, 336, 41], "type": "Text", "score": 0.94 } ]
48
+ ```
49
+
50
+ `box` is `[x, y, w, h]` in page pixels; `type` ∈ `Text`, `ChoiceButton`, `Signature`. Served as a single-file
51
+ ONNX graph (fp32 + fp16), verified to reproduce the benchmark (fp32 mAP 0.6776 / fp16 0.6743 over all variants).
52
+
53
+ ## Intended use and limits
54
+
55
+ - Empty, filled, and handwritten digital/scanned forms are all supported.
56
+ - GPU/server model (~57 fps on an A100). For on-device CPU use `form-field-v1-nano` / `-small`.
57
+ - Very dense pages benefit from higher input resolution.
58
+
59
+ ## License & data
60
+
61
+ Weights are offered under a commercial Nutrient license. Evaluation is public and reproducible via
62
+ [`form-field-v1-benchmark`](https://huggingface.co/datasets/nutrientdocs/form-field-v1-benchmark). The FFDetr
63
+ baseline is [FF-DETR / CommonForms](https://huggingface.co/jbarrow/FFDetr) by jbarrow, Apache-2.0.
64
+
65
+ > ### πŸ“© Get access
66
+ >
67
+ > `form-field-v1-medium` is commercial and its weights are not downloadable here. To run it on-prem, with
68
+ > documents staying in your infrastructure β€” **contact Nutrient:
69
+ > [nutrient.io/contact-sales](https://www.nutrient.io/contact-sales/).**
70
+
71
+ ## About the author
72
+
73
+ <a href="https://nutrient.io/">
74
+ <img src="https://avatars2.githubusercontent.com/u/1527679?v=3&s=200" height="80" />
75
+ </a>
76
+
77
+ This project is maintained and funded by [Nutrient](https://nutrient.io/) - The deterministic document infrastructure enterprises run their highest-stakes workflows on: replayable output, clear exceptions, and full audit trails on the messy, regulated documents where AI alone breaks.