add: usecase 2(Image describer, streamlit frontend)

app.py

@@ -1,37 +1,96 @@
+import os
+import tempfile
+
 import streamlit as st
 
 from src.pipeline.main import LearnableAI
 
 
 def main():
-    st.title("LearnableAI Interface(Use Case 1)")
+    st.title("LearnableAI")
 
     learnable_ai = LearnableAI()
 
-    words = st.text_input("Enter your words:", "")
-    interest = st.text_input("Enter your interest:", "")
-    difficulty = st.selectbox(
-        "Select difficulty level:",
-        ("Beginner", "Easy", "Challenging", "Expert",
-         "Starter", "Basic", "Tough", "Master")
+    st.sidebar.header("Configuration")
+    use_case = st.sidebar.radio(
+        "Select Use Case",
+        ["Use Case: 1(Word-Sentences)", "Use Case: 2(Image Describer)"]
+    )
+
+    difficulty = st.sidebar.selectbox(
+        "Select Difficulty Level",
+        ["Simple", "Easy", "Challenging", "Advanced"]
     )
 
-    if st.button("Process with LearnableAI"):
-        if words and interest:  # Only show warning if fields are empty
-            try:
-                result = learnable_ai.word_to_sentence(
-                    words=list(words.split(",")),
-                    interest=interest,
-                    difficulties=difficulty
-                )
+    if use_case == "Use Case: 1(Word-Sentences)":
+        st.header("Use Case: 2(Image Describer)")
+
+        words = st.text_area(
+            "Words:",
+            "",
+            help="Enter multiple words separated by commas"
+        )
+
+        interest = st.text_input("Child Interest:", "",
+                                 help="Enter the interest area for the content ie. Superheroes, Animals, etc.")
+
+        if st.button("Generate Content", key="text_process"):
+            if words and interest:
+                try:
+                    word_list = [w.strip() for w in words.split(",")] if "," in words else [words.strip()]
+                    result = learnable_ai.word_to_sentence(
+                        words=word_list,
+                        interest=interest,
+                        difficulties=difficulty
+                    )
+
+                    st.success("Content Generated Successfully!")
+                    st.write("### Results")
+                    st.write(result)
+
+                except Exception as e:
+                    st.error(f"An error occurred: {str(e)}")
+            else:
+                st.warning("Please fill in both words and interest fields!")
 
-                st.write("### Results")
-                st.write(result)
+    else:  # Image Description
+        st.header("Image Description")
 
-            except Exception as e:
-                st.error(f"An error occurred: {str(e)}")
+        upload_method = st.radio(
+            "Choose Upload Method",
+            ["Upload File", "Capture from Camera"]
+        )
+
+        if upload_method == "Upload File":
+            image_file = st.file_uploader(
+                "Upload an image",
+                type=["jpg", "jpeg", "png"]
+            )
         else:
-            st.warning("Please fill in both words and interest fields!")
+            image_file = st.camera_input("Take a picture")
+
+        if image_file is not None:
+            st.image(image_file, caption="Uploaded Image", use_column_width=True)
+
+            if st.button("Generate Description", key="image_process"):
+                try:  # temp image file
+                    with tempfile.NamedTemporaryFile(delete=False, suffix='.jpg') as temp_file:
+                        temp_file.write(image_file.read())
+                        temp_image_path = temp_file.name
+
+                    result = learnable_ai.image_describer(
+                        image_path=temp_image_path,
+                        difficulties=difficulty
+                    )
+
+                    st.success("Image Processed Successfully!")
+                    st.write("### Image Description Results")
+                    st.write(result)
+
+                    os.remove(temp_image_path)  # remove temp image file
+
+                except Exception as e:
+                    st.error(f"An error occurred while processing the image: {str(e)}")
 
 
 if __name__ == "__main__":

requirements.txt

@@ -1,2 +1,3 @@
 groq
-streamlit
+streamlit
+google-generativeai

src/pipeline/main.py

@@ -5,15 +5,24 @@ author  @ github.com/ishworrsubedii
 """
 import os
 
+import google.generativeai as genai
 from groq import Groq
 
 from src.services.use_case_one.word_to_sentence import UseCaseOne
+from src.services.use_case_two.image_describer import ImageDescriptionGenerator
 
 
 class LearnableAI:
     def __init__(self):
         self.groq_client = Groq(api_key=os.getenv('GROQ_API_KEY'))
+        self.genai = genai.configure(api_key=os.getenv('GENAI_API_KEY'))
+
         self.use_case_one = UseCaseOne(client=self.groq_client)
 
+        self.use_case_two = ImageDescriptionGenerator()
+
     def word_to_sentence(self, interest: str, difficulties: str, words: list):
         return self.use_case_one.generate_educational_content(interest, difficulties, words)
+
+    def image_describer(self, image_path: str, difficulties):
+        return self.use_case_two.process_image(image_path, difficulty_level=difficulties)

src/services/use_case_two/image_describer.py

@@ -0,0 +1,63 @@
+"""
+project @ LearnableAI
+created @ 2025-01-17
+author  @ github.com/ishworrsubedii
+"""
+import google.generativeai as genai
+from PIL import Image
+
+
+class ImageDescriptionGenerator:
+    def __init__(self):
+        self.model = genai.GenerativeModel('gemini-1.5-flash')
+
+    def process_image(self, image_path, difficulty_level, ):
+        try:
+            img = Image.open(image_path)
+
+            prompt = self._create_analysis_prompt(difficulty_level)
+
+            response = self.model.generate_content([prompt, img])
+
+            return response.text
+
+        except Exception as e:
+            return f"Error processing image: {str(e)}"
+
+    def _create_analysis_prompt(self, difficulty_level):
+        return f"""
+        # Image Analysis for Children
+
+        Please analyze this image and create child-friendly content based on these parameters:
+        - Difficulty Level: {difficulty_level}
+
+        Generate the following:
+        1. Simple description of the image (2-3 sentences)
+        2. Three interesting details about what you see
+        3. Four age-appropriate questions about the image
+        4. One interactive activity related to the image
+
+        Requirements:
+        - Use simple, clear language appropriate for {difficulty_level} level
+        - Questions should encourage observation and critical thinking
+        - Include both easy and slightly challenging questions
+        - Activity should be engaging and educational
+
+        Format the response exactly like this:
+        DESCRIPTION:
+        [Image description]
+
+        INTERESTING DETAILS:
+        1. [Detail 1]
+        2. [Detail 2]
+        3. [Detail 3]
+
+        QUESTIONS:
+        Q1: [Question 1]
+        Q2: [Question 2]
+        Q3: [Question 3]
+        Q4: [Question 4]
+
+        ACTIVITY:
+        [Activity description]
+        """