palletiavi23 commited on
Commit
267cf00
·
verified ·
1 Parent(s): 83fc7ad

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +63 -1
README.md CHANGED
@@ -1,3 +1,65 @@
1
  ---
2
  license: other
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: other
3
+ Author: Intel Corporation
4
+ tags:
5
+ - patient
6
+ - reference-design
7
+ - evaluation-only
8
+ ---
9
+
10
+ ## Important Notice
11
+ This model is provided for evaluation and development purposes
12
+ only. It is not validated for and must not be used in clinical,
13
+ diagnostic, or production settings. See #use-and-limitations.
14
+
15
+ ---
16
+ ## Model Information
17
+
18
+ ### Description
19
+ patient-present is a single-stage object detection model that localizes
20
+ a neonate/infant patient inside a NICU radiant warmer from a video frame.
21
+ It uses a MobileNetV2 backbone with a Feature Pyramid Network (FPN) neck
22
+ and an anchor-free single-stage detection head, following the
23
+ mmdetection framework conventions. This model is a development and evaluation
24
+ tool produced as part of Intel's NICU Warmer reference design and has not
25
+ undergone clinical validation.
26
+
27
+ ### Intended Use
28
+ This model is intended for use by software developers and researchers evaluating AI-assisted monitoring workflows on Intel hardware. It is designed to demonstrate detection feasibility in a NICU warmer context. patient_present itself does not provide any medical functionality, nor is it intended to process or interpret medical data for a medical purpose. Developers are responsible for independently validating and adapting patient-present for their specific use case. It must not be used in live clinical environments, for diagnostic decisions, or on live patients.
29
+
30
+ ---
31
+ ## Technical Specifications
32
+
33
+ | Attribute | Detail |
34
+ |-------------------|-------------------------------------|
35
+ | Architecture | MobileNetV2 backbone + FPN neck + anchor-free detection head |
36
+ | Parameters | ~5–7M (FP32) |
37
+ | Input | 992×800 RGB image (W×H), float32, normalized [0,1], NCHW |
38
+ | Output | [N, 5] — bounding boxes (x1, y1, x2, y2, confidence), post-NMS, pixel coords at input resolution |
39
+ | Training hardware | Intel Ultra Core |
40
+ | Framework | PyTorch (mmdetection) → OpenVINO IR FP32 |
41
+
42
+ ---
43
+ ## Use and Limitations
44
+
45
+ ### Permitted Uses
46
+ - Evaluation and benchmarking of [domain] AI workflows
47
+ - Research and development
48
+ - Academic study
49
+ - Adaptation and modification for non-clinical applications
50
+
51
+ ### Prohibited Uses
52
+ - Clinical or diagnostic use
53
+ - Production deployment without independent validation and
54
+ applicable regulatory authorization
55
+ - Use with live patients or real patient data (unless the user
56
+ has obtained all necessary authorizations)
57
+ - Any use that would cause Intel to be treated as the
58
+ manufacturer, provider, or deployer of a medical device
59
+ or regulated AI system
60
+
61
+ ---
62
+ ## License
63
+ The use of people-present is governed by the Intel Non-Clinical
64
+ Evaluation Use License. By accessing this model, you
65
+ agree to the license terms.