SpaceYL commited on
Commit
5e6ddd0
·
verified ·
1 Parent(s): 719380e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +60 -54
README.md CHANGED
@@ -1,72 +1,78 @@
1
  ---
2
- language:
3
- - en
4
- - fr
5
  license: apache-2.0
6
- datasets:
7
- - UMA-IA/UMA_Dataset_Engine_Aero_VLM
8
- base_model: Qwen/Qwen2.5-VL-7B-Instruct
9
- tags:
10
- - aerospace
11
- - aeronautics
12
- - engineering
13
- - vision-language
14
- - component-detection
15
- pipeline_tag: image-to-text
16
  ---
17
 
18
- ## Model Details
19
 
20
- **Model Name:** UMA-IA/VLM_Engine_Finetuned_Aero
21
- **Authors:**
22
  - **Youri LALAIN**, Engineering student at French Engineering School ECE
23
  - **Lilian RAGE**, Engineering student at French Engineering School ECE
24
 
25
- **Base Model:** [Qwen/Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct)
26
- **Fine-tuned Dataset:** [UMA-IA/UMA_Dataset_Engine_Aero_VLM](https://huggingface.co/datasets/UMA-IA/UMA_Dataset_Engine_Aero_VLM)
27
- **License:** Apache 2.0
28
 
29
- ## Model Description
30
 
31
- # Qwen2.5-VL Fine-tuned pour la détection de composants de moteurs aérospatiaux
32
 
33
- VLM_Engine_Finetuned_Aero est une version spécialisée du modèle Qwen2.5-VL-7B-Instruct, fine-tunée pour détecter, identifier et analyser les composants de moteurs aéronautiques et aérospatiaux à partir d'images. Le modèle exploite le dataset UMA-IA/UMA_Dataset_Engine_Aero_VLM pour améliorer sa capacité à reconnaître les pièces spécifiques, les défauts potentiels et les caractéristiques techniques des systèmes de propulsion.
 
 
 
 
 
 
 
 
 
34
 
35
- ## Capacités
36
- - Détection et identification précise des composants de moteurs aéronautiques
37
- - Analyse visuelle des pièces mécaniques et de leur état
38
- - Reconnaissance des défauts ou anomalies sur les composants
39
- - Fourniture d'informations techniques sur les pièces identifiées
40
- - Assistance au diagnostic visuel pour la maintenance
41
 
42
- ## Cas d'utilisation
43
- - Formation des techniciens et ingénieurs aéronautiques
44
- - Assistance à la documentation technique
45
- - Aide visuelle
46
 
47
- ## Détails de l'entraînement
48
- Ce modèle a été fine-tuné sur UMA-IA/UMA_Dataset_Engine_Aero_VLM, un dataset spécialement créé pour l'identification visuelle de composants de moteurs aérospatiaux. L'entraînement a été réalisé en utilisant des techniques de fine-tuning supervisé pour adapter le modèle Qwen2.5-VL à la reconnaissance de composants techniques spécifiques.
 
 
 
 
 
 
49
 
50
- ## Comment utiliser le modèle
51
- Vous pouvez charger le modèle en utilisant la bibliothèque `transformers` de Hugging Face :
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52
 
53
  ```python
54
- from transformers import AutoModelForCausalLM, AutoTokenizer
 
 
 
 
 
 
55
  from PIL import Image
56
- import requests
57
- from io import BytesIO
58
-
59
- # Charger le modèle et le tokenizer
60
- model_name = "UMA-IA/VLM_Engine_Finetuned_Aero"
61
- model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True)
62
- tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
63
-
64
- # Charger une image (exemple avec une URL)
65
- image_url = "URL_DE_VOTRE_IMAGE"
66
- response = requests.get(image_url)
67
- image = Image.open(BytesIO(response.content))
68
-
69
- # Préparer la requête
70
- prompt = "Identifiez les composants visibles dans cette image de moteur d'avion et décrivez leur fonction."
71
- response = model.chat(tokenizer, query=prompt, image=image)
72
- print(response)
 
1
  ---
 
 
 
2
  license: apache-2.0
 
 
 
 
 
 
 
 
 
 
3
  ---
4
 
5
+ # UMA-IA/UMA_Dataset_Engine_Aero_VLM
6
 
7
+ ## Authors
 
8
  - **Youri LALAIN**, Engineering student at French Engineering School ECE
9
  - **Lilian RAGE**, Engineering student at French Engineering School ECE
10
 
11
+ ## Dataset Summary
 
 
12
 
13
+ The **UMA-IA/UMA_Dataset_Engine_Aero_VLM** is a specialized dataset designed for training vision-language models in the field of **aerospace and aeronautical engineering**. It consists of high-quality **images of aircraft engine components paired with detailed captions** identifying and describing the visible parts. This dataset enables models to learn to recognize and analyze various engine components, making it ideal for **fine-tuning vision-language models** for technical visual recognition and analysis tasks in the aerospace industry.
14
 
15
+ ## Dataset Details
16
 
17
+ - **Total Samples**: Comprehensive collection of aerospace engine component images
18
+ - **Splits**:
19
+ - **Train**: Primary training set for model development
20
+ - **Validation**: For model evaluation during development
21
+ - **Test**: For final performance assessment
22
+ - **Columns**:
23
+ - `image`: The image file of an aircraft engine component or cross-section
24
+ - `caption`: Detailed description of visible components in the image
25
+ - `image_id`: Unique identifier for each image
26
+ - `cui`: Technical classification identifier
27
 
28
+ ## Dataset Structure
 
 
 
 
 
29
 
30
+ The dataset's primary focus is on providing high-quality annotated images of aircraft engine components with detailed technical descriptions. Each entry contains:
 
 
 
31
 
32
+ 1. An image showing aerospace engine components from various angles and cross-sections
33
+ 2. Detailed captions identifying components such as:
34
+ - Soufflante (Fan)
35
+ - Aubes (Blades)
36
+ - Rotor
37
+ - Stator
38
+ - Compresseur (Compressor)
39
+ - And other critical engine components
40
 
41
+ ## Example Entries
42
+
43
+ | image | caption | image_id |
44
+ |-------|---------|----------|
45
+ | [Engine Image] | Composants visibles: - Soufflante - Aubes de soufflante - Rotor de soufflante... | 001269777_896x598_c_mirror |
46
+ | [Engine Image] | Composants visibles: - Soufflante - Aubes - Rotor - Stator - Compresseur... | 001269777_896x598_c_original |
47
+ | [Engine Image] | Composants visibles: - Soufflante - Aubes - Rotor - Stator - Compresseur... | 001269777_896x598_c_segment1 |
48
+
49
+ ## Applications
50
+
51
+ This dataset is particularly valuable for:
52
+ - Training vision models to recognize aerospace engine components
53
+ - Developing diagnostic tools for engine maintenance
54
+ - Creating educational resources for aerospace engineering
55
+ - Enhancing technical documentation with automatic component recognition
56
+ - Supporting quality control processes in engine manufacturing
57
+
58
+ ## How to Use
59
+
60
+ You can load this dataset using the Hugging Face `datasets` library:
61
 
62
  ```python
63
+ from datasets import load_dataset
64
+ dataset = load_dataset("UMA-IA/UMA_Dataset_Engine_Aero_VLM")
65
+
66
+ # Access the first sample
67
+ print(dataset["train"][0]["caption"])
68
+
69
+ # Display an image (if in a notebook environment)
70
  from PIL import Image
71
+ import matplotlib.pyplot as plt
72
+
73
+ img = dataset["train"][0]["image"]
74
+ plt.figure(figsize=(10, 8))
75
+ plt.imshow(img)
76
+ plt.axis('off')
77
+ plt.title(dataset["train"][0]["caption"])
78
+ plt.show()