You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: articles/applied-ai-services/form-recognizer/concept-composed-models.md
+5-5Lines changed: 5 additions & 5 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -34,15 +34,15 @@ With composed models, you can assign multiple custom models to a composed model
34
34
35
35
* For **_custom template models_**, the composed model can be created using variations of a custom template or different form types. This operation is useful when incoming forms may belong to one of several templates.
36
36
37
-
* The response will include a ```docType``` property to indicate which of the composed models was used to analyze the document.
37
+
* The response includes a ```docType``` property to indicate which of the composed models was used to analyze the document.
38
38
39
39
* For ```Custom neural``` models the best practice is to add all the different variations of a single document type into a single training dataset and train on custom neural model. Model compose is best suited for scenarios when you have documents of different types being submitted for analysis.
40
40
41
41
::: moniker-end
42
42
43
43
::: moniker range="form-recog-3.0.0"
44
44
45
-
With the introduction of [****custom classifier models****](./concept-custom-classifier.md), you have the option to continue to use [**composed models**](./concept-composed-models.md) or the classifier model as an explicit step before analysis. For a deeper understanding of when to use a classifier or composed model, _see_[**Custom classifier models**](concept-custom-classifier.md).
45
+
With the introduction of [****custom classifier models****](./concept-custom-classifier.md), you can choose to use [**composed models**](./concept-composed-models.md) or the classifier model as an explicit step before analysis. For a deeper understanding of when to use a classifier or composed model, _see_[**Custom classifier models**](concept-custom-classifier.md).
46
46
::: moniker-end
47
47
48
48
## Compose model limits
@@ -60,9 +60,9 @@ With the introduction of [****custom classifier models****](./concept-custom-cla
* To compose a model trained with a prior version of the API (v2.1 or earlier), train a model with the v3.0 API using the same labeled dataset. That addition will ensure that the v2.1 model can be composed with other models.
63
+
* To compose a model trained with a prior version of the API (v2.1 or earlier), train a model with the v3.0 API using the same labeled dataset. That addition ensures that the v2.1 model can be composed with other models.
64
64
65
-
* Models composed with v2.1 of the API will continue to be supported, requiring no updates.
65
+
* Models composed with v2.1 of the API continues to be supported, requiring no updates.
66
66
67
67
* The limit for maximum number of custom models that can be composed is 100.
68
68
@@ -81,7 +81,7 @@ The following resources are supported by Form Recognizer **v3.0** :
81
81
82
82
::: moniker range="form-recog-2.1.0"
83
83
84
-
The following resources are supported by Form Recognizer v2.1:
84
+
Form Recognizer v2.1 supports the following resources:
Custom classifier models are deep-learning-model types that combines layout and language features to accurately detect and identify documents you process within your application. Custom classifier models can classify each page in a input file to identify the document(s) within and can also identify multiple documents or multiple instances of a single document within an input file.
21
+
Custom classifier models are deep-learning-model types that combine layout and language features to accurately detect and identify documents you process within your application. Custom classifier models can classify each page in an input file to identify the document(s) within and can also identify multiple documents or multiple instances of a single document within an input file.
22
22
23
23
## Model capabilities
24
24
@@ -30,7 +30,7 @@ Custom classifier models can analyze a single- or multi-file documents to identi
30
30
31
31
* A single file containing multiple instances of the same document. For instance, a collection of scanned invoices.
32
32
33
-
Training a custom classifier model requires at least 2 distinct classes and a minimum of 5 samples per class.
33
+
Training a custom classifier model requires at least two distinct classes and a minimum of five samples per class.
34
34
35
35
### Compare custom classifier and composed models
36
36
@@ -39,22 +39,22 @@ A custom classifier model can replace [a composed model](concept-composed-models
39
39
| Capability | Custom classifier process | Composed model process |
40
40
|--|--|--|
41
41
|Analyze a single document of unknown type belonging to one of the types trained for extraction model processing.|● Requires multiple calls. </br> ● Call the classifier models based on the document class. This step allows for a confidence-based check before invoking the extraction model analysis.</br> ● Invoke the extraction model. |● Requires a single call to a composed model containing the model corresponding to the input document type. |
42
-
|Analyze a single document of unknown type belonging to several types trained for extraction model processing.|●Requires multiple calls.</br> ● Make a call to the classifier that ignores documents not matching a designated type for extraction.</br> ● Invoke the extraction model. |● Requires a single call to a composed model. The service will always pick a custom model within the composed model with the highest match.</br> ● A composed model cannot ignore documents.|
43
-
|Analyze a file containing multiple documents of known or unknown type belonging to one of the types trained for extraction model processing.|● Requires multiple calls. </br> ● Call the extraction model for each identified document in the input file.</br> ● Invoke the extraction model. |● Requires a single call to a composed model.</br> ● The composed model will only invoke the component model once on the first instance of the document. </br> ●The remaining documents are ignored. |
42
+
|Analyze a single document of unknown type belonging to several types trained for extraction model processing.|●Requires multiple calls.</br> ● Make a call to the classifier that ignores documents not matching a designated type for extraction.</br> ● Invoke the extraction model. |● Requires a single call to a composed model. The service selects a custom model within the composed model with the highest match.</br> ● A composed model can't ignore documents.|
43
+
|Analyze a file containing multiple documents of known or unknown type belonging to one of the types trained for extraction model processing.|● Requires multiple calls. </br> ● Call the extraction model for each identified document in the input file.</br> ● Invoke the extraction model. |● Requires a single call to a composed model.</br> ● The composed model invokes the component model once on the first instance of the document. </br> ●The remaining documents are ignored. |
44
44
45
45
## Language support
46
46
47
47
Classifier models currently only support English language documents.
48
48
49
49
## Best practices
50
50
51
-
Custom classifier models require a minimum of five samples per class to train. If the classes are very similar, adding additional training samples will improve model accuracy.
51
+
Custom classifier models require a minimum of five samples per class to train. If the classes are similar, adding extra training samples improves model accuracy.
52
52
53
53
## Training a model
54
54
55
55
Custom classifier models are only available in the [v3.0 API](v3-migration-guide.md) starting with API version ```2023-02-28-preview```. [Form Recognizer Studio](https://formrecognizer.appliedai.azure.com/studio) provides a no-code user interface to interactively train a custom classifier.
56
56
57
-
When using the REST API, if your documents are organized by folders, you can use the ```azureBlobSource``` property of the request to train a classifier model.
57
+
When using the REST API, if you've organized your documents by folders, you can use the ```azureBlobSource``` property of the request to train a classifier model.
Alternatively, if you have a flat list of files or only plan to use a few select files within each folder to train the model, you can use the ```azureBlobFileListSource``` property to train the model. This requires an additional```file list``` in [JSON Lines](https://jsonlines.org/) format. For each class, add a new file with a list of files to be submitted for training.
89
+
Alternatively, if you have a flat list of files or only plan to use a few select files within each folder to train the model, you can use the ```azureBlobFileListSource``` property to train the model. This step requires a```file list``` in [JSON Lines](https://jsonlines.org/) format. For each class, add a new file with a list of files to be submitted for training.
Copy file name to clipboardExpand all lines: articles/applied-ai-services/form-recognizer/concept-custom-neural.md
+6-6Lines changed: 6 additions & 6 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -30,13 +30,13 @@ Custom neural models share the same labeling format and strategy as [custom temp
30
30
31
31
## Model capabilities
32
32
33
-
Custom neural models currently only support key-value pairs and selection marks and structured fields (tables), future releases will include support for signatures.
33
+
Custom neural models currently only support key-value pairs and selection marks and structured fields (tables), future releases include support for signatures.
34
34
35
35
| Form fields | Selection marks | Tabular fields | Signature | Region |
<sup>1</sup> Region labels in custom neural models will use the results from the Layout API for specified region. This feature is different from template models where, if no value is present, text is generated at training time.
39
+
<sup>1</sup> Region labels in custom neural models use the results from the Layout API for specified region. This feature is different from template models where, if no value is present, text is generated at training time.
40
40
41
41
### Build mode
42
42
@@ -46,11 +46,11 @@ Neural models support documents that have the same information, but different pa
46
46
47
47
## Language support
48
48
49
-
Neural models now support additional languages in the ```2023-02-28-preview``` API.
49
+
1.Neural models now support added languages in the ```2023-02-28-preview``` API.
50
50
51
51
| Languages | API version |
52
52
|:--:|:--:|
53
-
| English |`2022-08-31` (GA), `2023-02-28-preview`|
53
+
| English |`2022-08-31` (GA), `2023-02-28-preview`|
54
54
| German |`2023-02-28-preview`|
55
55
| Italian |`2023-02-28-preview`|
56
56
| French |`2023-02-28-preview`|
@@ -111,7 +111,7 @@ Custom neural models can generalize across different formats of a single documen
111
111
112
112
### Field naming
113
113
114
-
When you label the data, labeling the field relevant to the value will improve the accuracy of the key-value pairs extracted. For example, for a field value containing the supplier ID, consider naming the field "supplier_id". Field names should be in the language of the document.
114
+
When you label the data, labeling the field relevant to the value improves the accuracy of the key-value pairs extracted. For example, for a field value containing the supplier ID, consider naming the field "supplier_id". Field names should be in the language of the document.
115
115
116
116
### Labeling contiguous values
117
117
@@ -127,7 +127,7 @@ Values in training cases should be diverse and representative. For example, if a
127
127
## Current Limitations
128
128
129
129
* The model doesn't recognize values split across page boundaries.
130
-
* Custom neural models are only trained in English and model performance will be lower for documents in other languages.
130
+
* Custom neural models are only trained in English. Model performance is lower for documents in other languages.
131
131
* If a dataset labeled for custom template models is used to train a custom neural model, the unsupported field types are ignored.
132
132
* Custom neural models are limited to 10 build operations per month. Open a support request if you need the limit increased.
Copy file name to clipboardExpand all lines: articles/applied-ai-services/form-recognizer/concept-custom.md
+7-7Lines changed: 7 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -38,9 +38,9 @@ To create a custom extraction model, label a dataset of documents with the value
38
38
39
39
### Custom template model
40
40
41
-
The custom template or custom form model relies on a consistent visual template to extract the labeled data. The accuracy of your model is affected by variances in the visual structure of your documents. Structured forms such as questionnaires or applications are examples of consistent visual templates.
41
+
The custom template or custom form model relies on a consistent visual template to extract the labeled data. Variances in the visual structure of your documents affect the accuracy of your model. Structured forms such as questionnaires or applications are examples of consistent visual templates.
42
42
43
-
Your training set will consist of structured documents where the formatting and layout are static and constant from one document instance to the next. Custom template models support key-value pairs, selection marks, tables, signature fields, and regions. Template models and can be trained on documents in any of the [supported languages](language-support.md). For more information, *see*[custom template models](concept-custom-template.md).
43
+
Your training set consists of structured documents where the formatting and layout are static and constant from one document instance to the next. Custom template models support key-value pairs, selection marks, tables, signature fields, and regions. Template models and can be trained on documents in any of the [supported languages](language-support.md). For more information, *see*[custom template models](concept-custom-template.md).
44
44
45
45
> [!TIP]
46
46
>
@@ -71,7 +71,7 @@ This table provides links to the build mode programming language SDK references
71
71
72
72
## Compare model features
73
73
74
-
The table below compares custom template and custom neural features:
74
+
The following table compares custom template and custom neural features:
@@ -83,11 +83,11 @@ The table below compares custom template and custom neural features:
83
83
84
84
### Custom classifier model
85
85
86
-
Document classification is a new scenario supported by Form Recognizer with the ```2023-02-28-preview``` API. Document classifier supports classification and splitting scenarios. Train a classifier model to identify the different types of documents your application supports. The input file for the classifier model can contain multiple documents, the classifier will classify each document within an associated page range. See [custom classification](concept-custom-classifier.md) models to learn more.
86
+
Document classification is a new scenario supported by Form Recognizer with the ```2023-02-28-preview``` API. Document classifier supports classification and splitting scenarios. Train a classifier model to identify the different types of documents your application supports. The input file for the classifier model can contain multiple documents and classifies each document within an associated page range. See [custom classification](concept-custom-classifier.md) models to learn more.
87
87
88
88
## Custom model tools
89
89
90
-
The following tools are supported by Form Recognizer v3.0:
90
+
Form Recognizer v3.0 supports the following tools:
91
91
92
92
| Feature | Resources | Model ID|
93
93
|---|---|:---|
@@ -97,7 +97,7 @@ The following tools are supported by Form Recognizer v3.0:
97
97
98
98
::: moniker range="form-recog-2.1.0"
99
99
100
-
The following tools are supported by Form Recognizer v2.1:
100
+
Form Recognizer v2.1 supports the following tools:
101
101
102
102
> [!NOTE]
103
103
> Custom model types [custom neural](concept-custom-neural.md) and [custom template](concept-custom-template.md) are only available with Form Recognizer version v3.0.
@@ -205,7 +205,7 @@ This table compares the supported data extraction areas:
205
205
**Table symbols**:
206
206
✔—supported;
207
207
**n/a—currently unavailable;
208
-
*-behaves differently. With template models, synthetic data is generated at training time. With neural models, exiting text recognized in he region is selected.
208
+
*-behaves differently. With template models, synthetic data is generated at training time. With neural models, exiting text recognized in the region is selected.
209
209
210
210
> [!TIP]
211
211
> When choosing between the two model types, start with a custom neural model if it meets your functional needs. See [custom neural](concept-custom-neural.md) to learn more about custom neural models.
0 commit comments