Restructure project

app.py

@@ -0,0 +1,53 @@
+import streamlit as st
+
+st.set_page_config(
+    page_title="OpenEyes",
+    page_icon=":eyes:",
+)
+
+st.title("👀 OpenEyes")
+
+# Welcome message
+st.write("Welcome to the 👀 OpenEyes Homepage!")
+
+# goto the chat section to talk to the bot
+# st.write("Go to the chat section to talk to the bot!")
+st.info(
+    """👈 Go to the chat section to talk to the bot!
+"""
+)
+
+with st.expander("💡 Idea"):
+    st.markdown(
+        """
+        - The idea behind this project is to create a chatbot that can answer questions related to animals.
+        - The chatbot is trained on the [AnimalQA dataset](
+        https://www.kaggle.com/datasets/iamsouravbanerjee/animal-image-dataset-90-different-animals
+        ) from Kaggle.
+        We thought that this app will help the tourists who are visiting safari parks and zoos and want to know more about the animals they see.
+        They can scan the animals from their phones and our Object Detection model will detect the animal and then the chatbot will answer the questions related to the animal.
+        """
+    )
+
+with st.expander("🐍 Python Libraries Used"):
+    st.markdown(
+        """
+        - `Streamlit` - For the frontend
+        - `OpenAI` - For the chatbot
+        - `Requests` - For making HTTP requests
+        - `Time` - For adding delays
+        - `Base64` - For encoding and decoding
+        - `JSON` - For parsing JSON
+        - `Pillow` - For image processing
+        - `Numpy` - For numerical processing
+        - `Pandas` - For data processing
+        - `Matplotlib` - For data visualization
+        - `YOLOv8` - For object detection
+        - `PyTorch` - For deep learning
+        - `Transformers` - For NLP
+        - `HuggingFace` - For NLP
+        - `Whisper` - For speech recognition
+        - `PyAudio` - For audio processing
+        
+        """
+    )

packages.txt

@@ -0,0 +1 @@
+libgl1

pages/.cph/.yolo.py_ab91d06c1049f91607ef5865ba1b5ac0.prob

@@ -0,0 +1 @@
+{"name":"Local: yolo","url":"c:\\Users\\hp\\Documents\\GitHub\\OpenEyes\\Streamlit\\pages\\yolo.py","tests":[{"id":1688839560533,"input":"","output":""}],"interactive":false,"memoryLimit":1024,"timeLimit":3000,"srcPath":"c:\\Users\\hp\\Documents\\GitHub\\OpenEyes\\Streamlit\\pages\\yolo.py","group":"local","local":true}

pages/1_📷_Camera.py

@@ -0,0 +1,45 @@
+import streamlit as st
+from PIL import Image
+from ultralytics import YOLO
+
+# initialize the model
+model = YOLO("yolov8n.pt")
+
+# radio button to upload a picture or take a picture from webcam
+st.sidebar.subheader("Select Input Source")
+input_type = st.sidebar.radio(" ", ("Upload an Image", "Take a Picture"))
+
+global picture
+
+if input_type == "Upload an Image":
+    picture = st.file_uploader("Select an image", type=["jpg", "jpeg", "png"])
+else:
+    # display the picture
+    picture = st.camera_input("Take a picture")
+
+
+if picture:
+    st.image(picture, caption="Your picture", use_column_width=True)
+
+    # save the picture
+    with open("image.jpg", "wb") as f:
+        f.write(picture.getvalue())  # write the picture to the disk
+
+    st.info("Select the confidence threshold for the object detection model and press Detect Objects")
+
+    # confidence slider
+    st.sidebar.subheader("Confidence Threshold")
+    confidence_threshold = st.sidebar.slider("Select a value", 0.0, 1.0, 0.5, 0.01)
+
+    source_img = Image.open("image.jpg")
+
+    if st.sidebar.button("Detect Objects"):
+        # start the detection
+        res = model.predict(source_img, conf=confidence_threshold)
+        boxes = res[0].boxes
+        res_plotted = res[0].plot()[:, :, ::-1]
+        st.image(res_plotted, caption="Detected Image", use_column_width=True)
+
+        st.balloons()
+
+        st.success("Head over to the chat page to ask questions and link to the page")

pages/✉_Chat.py

@@ -0,0 +1,154 @@
+# import openai
+import streamlit as st
+import requests
+import time
+import base64
+import json
+
+st.set_page_config(
+    page_title="OpenEyes",
+    page_icon="👀",
+)
+
+st.title("OpenEyes")
+
+# API_KEY = st.secrets["OPENAI_API_KEY"]
+
+
+def makeRequest(prompt):
+    url = "http://3.88.181.187:8080/v1/"
+    headers = {"Content-Type": "application/json"}
+    data = {
+        "model": "gpt-4",
+        "messages": [{"role": "user", "content": prompt}],
+    }
+    response = requests.post(url, headers=headers, json=data)
+    return response.json()
+
+
+if "openai_model" not in st.session_state:
+    st.session_state["openai_model"] = "gpt-4"  # setting up a session state model
+
+if "messages" not in st.session_state:
+    # st.session_state.messages = [] # setting up a session state messages to store the messages
+    # set a default message hi to the bot before the user types anything
+    st.session_state.messages = [
+        {
+            "role": "assistant",
+            "content": "Hi, I'm OpenEyes. I'm here to help you with your queries related to Animals. What do you want to know?",
+        }
+    ]
+
+for message in st.session_state.messages:
+    with st.chat_message(message["role"]):  # creating a chat message with the role
+        st.markdown(message["content"])  # adding the content to the chat message
+
+
+from transformers import pipeline
+from transformers import (
+    AutoTokenizer,
+    AutoModelForSequenceClassification,
+    TrainingArguments,
+    Trainer,
+)
+
+
+class CommandDetector:
+    def __init__(self, model_path, tokenizer="bert-base-uncased"):
+        self.tokenizer = AutoTokenizer.from_pretrained("bert-base-uncased")
+        self.classifier = pipeline(
+            "text-classification", model=model_path, tokenizer=tokenizer
+        )
+
+    def command_filter(self, prompt):
+        # Classify the input prompt
+        result = self.classifier(prompt)
+        command_id = int(result[0]["label"].split("_")[-1])
+        command = {0: "vision", 1: "chat", 2: "goodbye"}[command_id]
+
+        return command
+
+
+# Accept user input
+if prompt := st.chat_input("What is up?"):
+    # Add user message to chat history
+    st.session_state.messages.append({"role": "user", "content": prompt})
+
+    # # Classify the input prompt as intent
+    # get the model path from ../../models/checkpoint-760
+    mycd = CommandDetector(model_path="./models/checkpoint-760")
+    # st.write(intent)
+    intent = mycd.command_filter(prompt)
+    # st.write(intent)
+    if intent == "vision":
+        # st.info("Head over to the camera page to take a picture 📷")
+        st.session_state.messages = []
+        st.session_state.messages.append(
+            {
+                "role": "assistant",
+                "content": "Head over to the camera page to take a picture 📷",
+            }
+        )
+        st.experimental_rerun()
+    elif intent == "goodbye":
+        st.session_state.messages = []
+        # st.info("Bye 👋")
+        st.session_state.messages.append(
+            {
+                "role": "assistant",
+                "content": "Bye 👋, loved talking to you. See you soon!",
+            }
+        )
+        st.experimental_rerun()
+    else:
+        # st.info("Head over to the chat page to ask questions and link to the page")
+        pass
+
+    # Display user message in chat message container
+    with st.chat_message("user"):
+        st.markdown(prompt)
+    # Display assistant response in chat message container
+    with st.chat_message("assistant"):
+        message_placeholder = st.empty()
+        full_response = ""
+        response = makeRequest(prompt)
+        for chunk in response["choices"][0]["message"]["content"].split():
+            full_response += chunk + " "
+            time.sleep(0.05)
+            # Add a blinking cursor to simulate typing
+            message_placeholder.markdown(full_response + "▌")
+        message_placeholder.markdown(full_response)
+        st.session_state.messages.append(
+            {"role": "assistant", "content": full_response}
+        )
+
+
+# 3 cols
+col1, col2, col3 = st.columns([1, 1, 1])
+
+with col1:
+    if st.button("Export Chat History"):
+        # Download the chat history as a json file
+        # Convert JSON data to a string and encode as UTF-8
+        json_data = json.dumps(st.session_state.messages).encode("utf-8")
+
+        b64 = base64.b64encode(json_data).decode()
+        href = f'<a href="data:application/json;base64,{b64}" download="example.json">Download JSON</a>'
+        st.markdown(href, unsafe_allow_html=True)
+
+with col3:
+    # Clear chat button
+    if st.button("Clear Chat"):
+        # remove all the messages from the session state
+        st.session_state.messages = []
+        # add a default message to the session state
+        st.session_state.messages.append(
+            {
+                "role": "assistant",
+                "content": "Hi, I'm OpenEyes. I'm here to help you with your queries related to Animals. What do you want to know?",
+            }
+        )
+        # rerun the app
+        st.experimental_rerun()
+
+# print(st.session_state.messages)

pages/🧑_About.py

@@ -0,0 +1,52 @@
+import streamlit as st
+
+
+st.set_page_config(
+    page_title="About OpenEyes Devs",
+    page_icon="👀",
+)
+
+TeamMember1 = {
+    "Name": "Adarsh Anand",
+    "image": "https://avatars.githubusercontent.com/u/73928744?v=4",
+    "Role": "Intern @Intel | ICPC'22 Regionalist | Specialist @CodeForces | Ex-Graphy | Knight @LeetCode | Google DSC Lead '22 | IIT Goa CSE '24",
+    "LinkedIn": "https://www.linkedin.com/in/adarsh-anand-iitgoa/",
+    "GitHub": "https://github.com/adarshanand67",
+    "description": "I am a Full-stack developer with a passion for problem-solving. I love working on backend and frontend projects, but my true love is React.js. I enjoy working on projects that are challenging and have the potential to make a positive impact on people's lives.",
+}
+
+TeamMember2 = {
+    "Name": "Aniket Akshay Chaudhri",
+    "image": "https://avatars.githubusercontent.com/u/79798301?v=4",
+    "Role": "Head @ Coding Club IIT Goa | Competitive Programmer | Web Developer | Android Developer | CSE @ IIT Goa",
+    "LinkedIn": "https://www.linkedin.com/in/aniketchaudhri/",
+    "GitHub": "https://github.com/AniketChaudhri/",
+    "description": "I am a CSE pre-final year Undergrad at IIT Goa. I have experience in various Tech Domains such as Web Development, App Development, Deep Learning.",
+}
+
+# adarsh = Image.open("/Streamlit/adarsh.jpg")
+# aniket = Image.open("/Streamlit/aniket.jpg")
+
+# layout 2 cols
+st.title("About OpenEyes")
+col1, col2, col3 = st.columns([3,1,3])
+
+# center text
+# st.markdown("<h1 style='text-align: center; color: black;'>Meet the Team</h1>", unsafe_allow_html=True)
+
+# create 2 cards showing the details
+
+def show_details(TeamMember):
+    st.image(TeamMember["image"], width=200)
+    st.header(TeamMember["Name"])
+    st.markdown(TeamMember["Role"])
+    st.subheader(f"[LinkedIn]({TeamMember['LinkedIn']})")
+    st.subheader(f"[GitHub]({TeamMember['GitHub']})")
+    st.write(TeamMember["description"])
+
+# add border
+with col1:
+    show_details(TeamMember2)
+
+with col3:
+    show_details(TeamMember1)