NoExitPrior commited on
Commit
2a9d63a
Β·
1 Parent(s): f5f56a5

Update README with custom Muse:Eye documentation

Browse files
Files changed (1) hide show
  1. README.md +112 -73
README.md CHANGED
@@ -1,73 +1,112 @@
1
- # Google AI Edge Gallery ✨
2
-
3
- [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE)
4
- [![GitHub release (latest by date)](https://img.shields.io/github/v/release/google-ai-edge/gallery)](https://github.com/google-ai-edge/gallery/releases)
5
-
6
- **Explore, Experience, and Evaluate the Future of On-Device Generative AI with Google AI Edge.**
7
-
8
- The Google AI Edge Gallery is an experimental app that puts the power of cutting-edge Generative AI models directly into your hands, running entirely on your Android *(available now)* and iOS *(coming soon)* devices. Dive into a world of creative and practical AI use cases, all running locally, without needing an internet connection once the model is loaded. Experiment with different models, chat, ask questions with images and audio clip, explore prompts, and more!
9
-
10
- Install the app today from Google Play
11
-
12
- <a href='https://play.google.com/store/apps/details?id=com.google.ai.edge.gallery'><img alt='Get it on Google Play' width="250" src='https://play.google.com/intl/en_us/badges/static/images/badges/en_badge_web_generic.png'/></a>
13
-
14
- For users without Google Play access, install the apk from the [**latest release**](https://github.com/google-ai-edge/gallery/releases/latest/)
15
-
16
- > [!IMPORTANT]
17
- > You must uninstall all previous versions of the app before installing this one. Past versions will no longer be working and supported.
18
-
19
- <img width="480" alt="01" src="https://github.com/user-attachments/assets/2a60c8d0-ef4e-4040-a948-fa73f6a622b4" />
20
- <img width="480" alt="02" src="https://github.com/user-attachments/assets/d155d458-b822-415d-9252-7e825fe8c9c0" />
21
- <img width="480" alt="03" src="https://github.com/user-attachments/assets/1977af6f-ee7e-41b3-aac1-a642c66c0058" />
22
- <img width="480" alt="04" src="https://github.com/user-attachments/assets/a48be969-f57e-4497-9ecf-8feb35f2ba71" />
23
- <img width="480" alt="05" src="https://github.com/user-attachments/assets/2a9679ea-f191-4ffd-87db-6726f7c1057d" />
24
-
25
- ## ✨ Core Features
26
-
27
- * **πŸ“± Run Locally, Fully Offline:** Experience the magic of GenAI without an internet connection. All processing happens directly on your device.
28
- * **πŸ€– Choose Your Model:** Easily switch between different models from Hugging Face and compare their performance.
29
- * **πŸ–ΌοΈ Ask Image:** Upload images and ask questions about them. Get descriptions, solve problems, or identify objects.
30
- * **πŸŽ™οΈ Audio Scribe:** Transcribe an uploaded or recorded audio clip into text or translate it into another language.
31
- * **✍️ Prompt Lab:** Summarize, rewrite, generate code, or use freeform prompts to explore single-turn LLM use cases.
32
- * **πŸ’¬ AI Chat:** Engage in multi-turn conversations.
33
- * **πŸ“Š Performance Insights:** Real-time benchmarks (TTFT, decode speed, latency).
34
- * **🧩 Bring Your Own Model:** Test your local LiteRT `.litermlm` models.
35
- * **πŸ”— Developer Resources:** Quick links to model cards and source code.
36
-
37
- ## 🏁 Get Started in Minutes!
38
-
39
- 1. **Check OS Requirement**: Android 12 and up
40
- 2. **Download the App:**
41
- - Install the app from [Google Play](https://play.google.com/store/apps/details?id=com.google.ai.edge.gallery).
42
- - For users without Google Play access: install the apk from the [**latest release**](https://github.com/google-ai-edge/gallery/releases/latest/)
43
- 3. **Install & Explore:** For detailed installation instructions (including for corporate devices) and a full user guide, head over to our [**Project Wiki**](https://github.com/google-ai-edge/gallery/wiki)!
44
-
45
- ## πŸ› οΈ Technology Highlights
46
-
47
- * **Google AI Edge:** Core APIs and tools for on-device ML.
48
- * **LiteRT:** Lightweight runtime for optimized model execution.
49
- * **LLM Inference API:** Powering on-device Large Language Models.
50
- * **Hugging Face Integration:** For model discovery and download.
51
-
52
- ## ⌨️ Development
53
-
54
- Check out the [development notes](DEVELOPMENT.md) for instructions about how to build the app locally.
55
-
56
- ## 🀝 Feedback
57
-
58
- This is an **experimental Beta release**, and your input is crucial!
59
-
60
- * 🐞 **Found a bug?** [Report it here!](https://github.com/google-ai-edge/gallery/issues/new?assignees=&labels=bug&template=bug_report.md&title=%5BBUG%5D)
61
- * πŸ’‘ **Have an idea?** [Suggest a feature!](https://github.com/google-ai-edge/gallery/issues/new?assignees=&labels=enhancement&template=feature_request.md&title=%5BFEATURE%5D)
62
-
63
- ## πŸ“„ License
64
-
65
- Licensed under the Apache License, Version 2.0. See the [LICENSE](LICENSE) file for details.
66
-
67
- ## πŸ”— Useful Links
68
-
69
- * [**Project Wiki (Detailed Guides)**](https://github.com/google-ai-edge/gallery/wiki)
70
- * [Hugging Face LiteRT Community](https://huggingface.co/litert-community)
71
- * [LLM Inference guide for Android](https://ai.google.dev/edge/mediapipe/solutions/genai/llm_inference/android)
72
- * [LiteRT-LM](https://github.com/google-ai-edge/LiteRT-LM)
73
- * [Google AI Edge Documentation](https://ai.google.dev/edge)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ title: "Muse:Eye β€” On-device Multimodal XR Docent"
3
+ emoji: 🎨
4
+ colorFrom: blue
5
+ colorTo: purple
6
+ sdk: other
7
+ pipeline_tag: other
8
+ tags:
9
+ - multimodal
10
+ - XR
11
+ - on-device
12
+ - tflite
13
+ - mobileclip
14
+ - android
15
+ - unity
16
+ - RAG
17
+ - museum-tech
18
+ ---
19
+
20
+ # Muse:Eye
21
+ **On-device Multimodal AI + XR Museum Docent for Android**
22
+
23
+ Muse:EyeλŠ” **카메라 기반 μž‘ν’ˆ 인식(MobileCLIP)**,
24
+ **μ˜¨λ””λ°”μ΄μŠ€ LLM(Gemma-3N)**,
25
+ **XR μΈν„°λž™μ…˜(Unity MR)**
26
+ 을 κ²°ν•©ν•œ ν•˜μ΄λΈŒλ¦¬λ“œ AI λ„μŠ¨νŠΈ μ‹œμŠ€ν…œμž…λ‹ˆλ‹€.
27
+
28
+ μ˜€ν”„λΌμΈ ν™˜κ²½μ—μ„œλ„ λΉ λ₯Έ 이미지 검색,
29
+ 메타데이터 기반 RAG,
30
+ 성인/어린이/μ „λ¬Έκ°€ λͺ¨λ“œ μ„€λͺ… 제곡이 κ°€λŠ₯ν•©λ‹ˆλ‹€.
31
+
32
+ ---
33
+
34
+ # ✨ Features
35
+
36
+ ### 1) ⚑ On-device AI (No Internet Required)
37
+ - MobileCLIP (TFLite) 이미지 μž„λ² λ”©
38
+ - μž‘ν’ˆ μœ μ‚¬λ„ 검색 (FAISS β†’ custom `.bin` index)
39
+ - Gemma-3N μ˜¨λ””λ°”μ΄μŠ€ LLM
40
+
41
+ ### 2) πŸ–Ό Artwork Recognition
42
+ - 카메라 인식 β†’ μ¦‰μ‹œ embedding μΆ”μΆœ
43
+ - RAG 기반 μž‘ν’ˆΒ·μž‘κ°€ μ„€λͺ… 생성
44
+
45
+ ### 3) πŸ‘ΆπŸ§‘β€πŸ¦³ Three Docent Modes
46
+ - 어린이 λͺ¨λ“œ
47
+ - 성인 λͺ¨λ“œ
48
+ - μ „λ¬Έκ°€ λͺ¨λ“œ
49
+
50
+ ### 4) πŸ₯½ XR Integrated
51
+ - Unity + Android Studio 톡합
52
+ - MR λͺ¨λ“œμ—μ„œ 3D UI 및 μŒμ„± μ•ˆλ‚΄ 제곡
53
+
54
+ ---
55
+ # πŸ— System Architecture
56
+
57
+ Muse:EyeλŠ” 두 κ°€μ§€ μ‹€ν–‰ λͺ¨λ“œλ₯Ό μ§€μ›ν•©λ‹ˆλ‹€:
58
+
59
+ - **Mobile Mode (On-device AI)** β€” μ™„μ „ μ˜€ν”„λΌμΈ
60
+ - **XR/MR Mode (Cloud Multimodal AI)** β€” Unity 기반 MR + Cloud API
61
+
62
+ ---
63
+
64
+ ## πŸ“± A. Mobile Mode (On-device AI)
65
+
66
+ ```plaintext
67
+ Mobile App (Android)
68
+ β”‚
69
+ β”œβ”€β”€ Camera Input
70
+ β”‚
71
+ β”œβ”€β”€ MobileCLIP (TFLite)
72
+ β”‚ └─ on-device embedding (512-d)
73
+ β”‚
74
+ β”œβ”€β”€ Embedding Index (.bin)
75
+ β”‚ └─ cosine similarity search (offline)
76
+ β”‚
77
+ β”œβ”€β”€ Local Metadata RAG
78
+ β”‚ └─ artwork metadata lookup
79
+ β”‚
80
+ β”œβ”€β”€ Gemma-3N On-device LLM
81
+ β”‚ └─ adult / child / expert explanation
82
+ β”‚
83
+ └── Android TTS
84
+ ```
85
+
86
+ ---
87
+
88
+ ## πŸ₯½ B. XR/MR Mode (Cloud Multimodal AI)
89
+
90
+ ```plaintext
91
+ Unity XR App (MR View)
92
+ β”‚
93
+ β”œβ”€β”€ MR Camera (RenderTexture)
94
+ β”‚
95
+ β”œβ”€β”€ JNI Bridge β†’ Android
96
+ β”‚ └─ frame bytes 전달
97
+ β”‚
98
+ β”œβ”€β”€ Cloud Multimodal API (Gemini)
99
+ β”‚ β”œβ”€ artwork analysis
100
+ β”‚ β”œβ”€ style / meaning extraction
101
+ β”‚ └─ multimodal reasoning
102
+ β”‚
103
+ β”œβ”€β”€ Cloud PromptManager
104
+ β”‚ β”œβ”€ 어린이 λͺ¨λ“œ
105
+ β”‚ β”œβ”€ 성인 λͺ¨λ“œ
106
+ β”‚ └─ μ „λ¬Έκ°€ λͺ¨λ“œ
107
+ β”‚
108
+ └── Unity 3D UI + Android TTS
109
+ β”œβ”€ floating info panel
110
+ β”œβ”€β”€ 3D guide elements
111
+ └── audio output
112
+ ```