heygen avatar app

.gitignore

@@ -1,4 +1,7 @@
 venv/
 __pycache__/
 output.mp4
-UPIBOT.mp4
+UPIBOT.mp4
+b4319fabebeb43e89b68f00f*
+b4319fabebeb43e89b68f00f584014f1_clips/
+uploaded_photo.*

app.py

@@ -1,68 +1,114 @@
 import streamlit as st
+import requests
+import time
+import os
+from moviepy.editor import VideoFileClip
+from heygen import upload_photo, create_video, get_video_status, download_video, list_talking_photos, split_video, AUDIO_URL
 from utils import *
 
+
+# Streamlit App Configuration
 st.set_page_config(
     page_title="Number to Video",
     page_icon="🎥",
-    initial_sidebar_state="collapsed",
+    initial_sidebar_state="expanded",
 )
 
-# Input number
-number = st.text_input("Enter a number between 1 and 99,99,999", "24")
-number = int(number)
+# Session State Initialization
+if "talking_photo_id" not in st.session_state:
+    st.subheader("Photo Requirements")
+    st.write("""
+    - The face should be intact and clearly visible.
+    - Preferably use real human faces.
+    - Ensure only one face is visible in the photo.
+    - Face resolution should be larger than 200x200 pixels.
+    - The photo should be in JPEG or PNG format.
+    """)
+    st.image("sample.jpg", use_container_width=True)
+    st.session_state.talking_photo_id = None
+
+if "video_id" not in st.session_state:
+    st.session_state.video_id = None
+
+if "video_url" not in st.session_state:
+    st.session_state.video_url = None
+
+# Sidebar
+st.sidebar.title("Settings")
+uploaded_photo = st.sidebar.file_uploader("Upload a photo", type=["jpeg", "png"])
+upload = st.sidebar.button("Upload Photo", use_container_width=True)
+
+if uploaded_photo and upload:
+    file_path = f"uploaded_photo.{uploaded_photo.type.split('/')[1]}"
+    with open(file_path, "wb") as f:
+        f.write(uploaded_photo.read())
+
+    file_type = "image/jpeg" if uploaded_photo.type in ["image/jpg", "image/jpeg"] else "image/png"
+    st.session_state.talking_photo_id = upload_photo(file_path, file_type)
+    st.info("Photo uploaded successfully.")
+    st.info("Now generate the video.")
+
+# Video Creation
+if st.session_state.talking_photo_id:
+    create_video_button = st.sidebar.button("Generate Video", use_container_width=True, key="generate_video")
+    if create_video_button:
+        title="Number to Video"
+        st.session_state.video_id = create_video(title, st.session_state.talking_photo_id, AUDIO_URL)
+        # st.sidebar.write(f"Video ID: {st.session_state.video_id}")
+
+# Video Status Check
+if st.session_state.video_id:
+    with st.spinner("Generating video... This may take 2-5 minutes."):
+        while True:
+            video_status = get_video_status(st.session_state.video_id)
+            st.sidebar.info(f"Video Status: {video_status['status']}")
+
+            if video_status["status"] == "completed":
+                st.session_state.video_url = video_status["video_url"]
+                st.sidebar.info("Video generation completed.")
+                break
+            elif video_status["status"] == "failed":
+                st.sidebar.warning("Video generation failed. Please try again.")
+                break
+
+            time.sleep(20)
+
+# Video Download and Display
+if st.session_state.video_url:
+    video_id = st.session_state.video_id
+    video_url = st.session_state.video_url
+    if not os.path.exists(f"{video_id}.mp4"):
+        download_video(video_id, video_url)
+        split_video(video_id)
+        video_path = f"{video_id}.mp4"
+    # download_video(video_id, video_url)
+    # split_video(video_id)
 
-if number:
+    # Input number
+    number = st.text_input("Enter a number between 1 and 99,99,999", "24")
+    number = int(number)
     col1, col2 = st.columns(2)
 
     with col1:
         with st.container(border=True):
-            st.subheader("Settings")
-            st.divider()
+            # st.subheader("Photo avatar")
+            # st.divider()
 
-            avatar = st.radio("Choose an avatar", ("Custom avatar", "Template avatar"))
-            avatar = "custom" if avatar == "Custom avatar" else "template"
+            # st.image("uploaded_photo.jpg", use_container_width=True)
 
             # Generate clip sequence based on the number
             clips = generate_clip_sequence(number)
             if clips:
                 st.write("Generated Clips Sequence:")
+                st.divider()
                 for i, clip in enumerate(clips):
                     st.write(f"Clip {i + 1}: {clip}")
-                submit = st.button("Create Video", use_container_width=True)
-
-    if clips:
-        with st.sidebar:
-            st.title("Advanced Settings")
-            trim_settings = []
-            for i, clip in enumerate(clips):
-                with st.container(border=True):
-                    st.write(f"Adjust settings for {clip}:")
-                    length = get_clip_duration(clip, avatar)
-                    default_trim_start = 0.10
-                    default_trim_end = length - 0.10
-
-                    # trim_bounds = st.slider(
-                    #     f"Trim bounds for {clip}", 
-                    #     0.0, length, 
-                    #     (default_trim_start, default_trim_end), 
-                    #     0.01,
-                    #     key=f"bounds_{i}"
-                    # )
-                    # trim_bounds = tuple([round(b, 2) for b in trim_bounds])
-                    # print(trim_bounds)
-                    # trim_settings.append(trim_bounds)
-
-                    trim_start = st.slider(f"Trim start for {clip}", 0.0, float(length), 0.00, 0.01, key=f"start_{i}")
-                    trim_end = st.slider(f"Trim end for {clip}", 0.0, float(length), 0.00, 0.01, key=f"end_{i}")
-                    trim_settings.append((trim_start, trim_end))
-            print(trim_settings)
+                submit = st.button("Create Video", use_container_width=True, key="create_video")
 
     with col2:
         with st.container(border=True):
             st.subheader("Output")
-            output_file = "output.mp4"
+            output_file = f"{video_id}_{number}_output.mp4"
             if submit:
-                pass
-                # Create the video with individual clip settings
-                create_advanced_video(number, trim_settings, avatar, output_file)
-                st.video(output_file, autoplay=True)
+                combine_video(number, video_id, output_file)
+                st.video(output_file, autoplay=True) 

app_v2.py

@@ -0,0 +1,68 @@
+import streamlit as st
+from utils import *
+
+st.set_page_config(
+    page_title="Number to Video",
+    page_icon="🎥",
+    initial_sidebar_state="collapsed",
+)
+
+# Input number
+number = st.text_input("Enter a number between 1 and 99,99,999", "24")
+number = int(number)
+
+if number:
+    col1, col2 = st.columns(2)
+
+    with col1:
+        with st.container(border=True):
+            st.subheader("Settings")
+            st.divider()
+
+            avatar = st.radio("Choose an avatar", ("Custom avatar", "Template avatar"))
+            avatar = "custom" if avatar == "Custom avatar" else "template"
+
+            # Generate clip sequence based on the number
+            clips = generate_clip_sequence(number)
+            if clips:
+                st.write("Generated Clips Sequence:")
+                for i, clip in enumerate(clips):
+                    st.write(f"Clip {i + 1}: {clip}")
+                submit = st.button("Create Video", use_container_width=True)
+
+    if clips:
+        with st.sidebar:
+            st.title("Advanced Settings")
+            trim_settings = []
+            for i, clip in enumerate(clips):
+                with st.container(border=True):
+                    st.write(f"Adjust settings for {clip}:")
+                    length = get_clip_duration(clip, avatar)
+                    default_trim_start = 0.10
+                    default_trim_end = length - 0.10
+
+                    # trim_bounds = st.slider(
+                    #     f"Trim bounds for {clip}", 
+                    #     0.0, length, 
+                    #     (default_trim_start, default_trim_end), 
+                    #     0.01,
+                    #     key=f"bounds_{i}"
+                    # )
+                    # trim_bounds = tuple([round(b, 2) for b in trim_bounds])
+                    # print(trim_bounds)
+                    # trim_settings.append(trim_bounds)
+
+                    trim_start = st.slider(f"Trim start for {clip}", 0.0, float(length), 0.00, 0.01, key=f"start_{i}")
+                    trim_end = st.slider(f"Trim end for {clip}", 0.0, float(length), 0.00, 0.01, key=f"end_{i}")
+                    trim_settings.append((trim_start, trim_end))
+            print(trim_settings)
+
+    with col2:
+        with st.container(border=True):
+            st.subheader("Output")
+            output_file = "output.mp4"
+            if submit:
+                pass
+                # Create the video with individual clip settings
+                create_advanced_video(number, trim_settings, avatar, output_file)
+                st.video(output_file, autoplay=True)

clips_metadata.json

@@ -2,194 +2,194 @@
     {
         "clip_number": 0,
         "start_time_ms": 0,
-        "end_time_ms": 376,
+        "end_time_ms": 356,
         "initial_length": "316",
         "file_name": "1_cut.mp4"
     },
     {
         "clip_number": 1,
-        "start_time_ms": 636,
-        "end_time_ms": 888,
+        "start_time_ms": 656,
+        "end_time_ms": 868,
         "file_name": "2_cut.mp4"
     },
     {
         "clip_number": 2,
-        "start_time_ms": 1179,
-        "end_time_ms": 1451,
+        "start_time_ms": 1199,
+        "end_time_ms": 1431,
         "file_name": "3_cut.mp4"
     },
     {
         "clip_number": 3,
-        "start_time_ms": 1762,
-        "end_time_ms": 2037,
+        "start_time_ms": 1782,
+        "end_time_ms": 2017,
         "file_name": "4_cut.mp4"
     },
     {
         "clip_number": 4,
-        "start_time_ms": 2373,
-        "end_time_ms": 2721,
+        "start_time_ms": 2393,
+        "end_time_ms": 2701,
         "file_name": "5_cut.mp4"
     },
     {
         "clip_number": 5,
-        "start_time_ms": 3007,
-        "end_time_ms": 3233,
+        "start_time_ms": 3027,
+        "end_time_ms": 3213,
         "file_name": "6_cut.mp4"
     },
     {
         "clip_number": 7,
-        "start_time_ms": 3743,
-        "end_time_ms": 4163,
+        "start_time_ms": 3763,
+        "end_time_ms": 4143,
         "file_name": "7_cut.mp4"
     },
     {
         "clip_number": 8,
-        "start_time_ms": 4390,
-        "end_time_ms": 4638,
+        "start_time_ms": 4410,
+        "end_time_ms": 4618,
         "file_name": "8_cut.mp4"
     },
     {
         "clip_number": 9,
-        "start_time_ms": 4956,
-        "end_time_ms": 5454,
+        "start_time_ms": 4976,
+        "end_time_ms": 5434,
         "file_name": "9_cut.mp4"
     },
     {
         "clip_number": 10,
-        "start_time_ms": 5712,
-        "end_time_ms": 6036,
+        "start_time_ms": 5732,
+        "end_time_ms": 6016,
         "file_name": "10_cut.mp4"
     },
     {
         "clip_number": 11,
-        "start_time_ms": 6309,
-        "end_time_ms": 6725,
+        "start_time_ms": 6329,
+        "end_time_ms": 6705,
         "file_name": "20_cut.mp4"
     },
     {
         "clip_number": 12,
-        "start_time_ms": 7020,
-        "end_time_ms": 7412,
+        "start_time_ms": 7040,
+        "end_time_ms": 7392,
         "file_name": "30_cut.mp4"
     },
     {
         "clip_number": 13,
-        "start_time_ms": 7749,
-        "end_time_ms": 8116,
+        "start_time_ms": 7769,
+        "end_time_ms": 8096,
         "file_name": "40_cut.mp4"
     },
     {
         "clip_number": 14,
-        "start_time_ms": 8443,
-        "end_time_ms": 8823,
+        "start_time_ms": 8463,
+        "end_time_ms": 8803,
         "file_name": "50_cut.mp4"
     },
     {
         "clip_number": 15,
-        "start_time_ms": 9107,
-        "end_time_ms": 9592,
+        "start_time_ms": 9127,
+        "end_time_ms": 9572,
         "file_name": "60_cut.mp4"
     },
     {
         "clip_number": 16,
-        "start_time_ms": 9891,
-        "end_time_ms": 10371,
+        "start_time_ms": 9911,
+        "end_time_ms": 10351,
         "file_name": "70_cut.mp4"
     },
     {
         "clip_number": 17,
-        "start_time_ms": 10607,
-        "end_time_ms": 10960,
+        "start_time_ms": 10627,
+        "end_time_ms": 10940,
         "file_name": "80_cut.mp4"
     },
     {
         "clip_number": 18,
-        "start_time_ms": 11182,
-        "end_time_ms": 11708,
+        "start_time_ms": 11202,
+        "end_time_ms": 11688,
         "file_name": "90_cut.mp4"
     },
     {
         "clip_number": 19,
-        "start_time_ms": 11938,
-        "end_time_ms": 12434,
+        "start_time_ms": 11958,
+        "end_time_ms": 12414,
         "file_name": "11_cut.mp4"
     },
     {
         "clip_number": 20,
-        "start_time_ms": 12722,
-        "end_time_ms": 13089,
+        "start_time_ms": 12742,
+        "end_time_ms": 13069,
         "file_name": "12_cut.mp4"
     },
     {
         "clip_number": 21,
-        "start_time_ms": 13394,
-        "end_time_ms": 13865,
+        "start_time_ms": 13414,
+        "end_time_ms": 13845,
         "file_name": "13_cut.mp4"
     },
     {
         "clip_number": 22,
-        "start_time_ms": 14183,
-        "end_time_ms": 14621,
+        "start_time_ms": 14203,
+        "end_time_ms": 14601,
         "file_name": "14_cut.mp4"
     },
     {
         "clip_number": 23,
-        "start_time_ms": 14939,
-        "end_time_ms": 15411,
+        "start_time_ms": 14959,
+        "end_time_ms": 15391,
         "file_name": "15_cut.mp4"
     },
     {
         "clip_number": 24,
-        "start_time_ms": 15684,
-        "end_time_ms": 16280,
+        "start_time_ms": 15704,
+        "end_time_ms": 16260,
         "file_name": "16_cut.mp4"
     },
     {
         "clip_number": 25,
-        "start_time_ms": 16572,
-        "end_time_ms": 17153,
+        "start_time_ms": 16592,
+        "end_time_ms": 17133,
         "file_name": "17_cut.mp4"
     },
     {
         "clip_number": 26,
-        "start_time_ms": 17383,
-        "end_time_ms": 17827,
+        "start_time_ms": 17403,
+        "end_time_ms": 17807,
         "file_name": "18_cut.mp4"
     },
     {
         "clip_number": 27,
-        "start_time_ms": 18028,
-        "end_time_ms": 18624,
+        "start_time_ms": 18048,
+        "end_time_ms": 18604,
         "file_name": "19_cut.mp4"
     },
     {
         "clip_number": 28,
-        "start_time_ms": 18910,
-        "end_time_ms": 19301,
+        "start_time_ms": 18930,
+        "end_time_ms": 19281,
         "file_name": "100_cut.mp4"
     },
     {
         "clip_number": 29,
-        "start_time_ms": 19578,
-        "end_time_ms": 20046,
+        "start_time_ms": 19598,
+        "end_time_ms": 20026,
         "file_name": "1000_cut.mp4"
     },
     {
         "clip_number": 30,
-        "start_time_ms": 20268,
-        "end_time_ms": 20626,
+        "start_time_ms": 20288,
+        "end_time_ms": 20606,
         "file_name": "100000_cut.mp4"
     },
     {
         "clip_number": 31,
-        "start_time_ms": 21018,
-        "end_time_ms": 22227,
+        "start_time_ms": 21038,
+        "end_time_ms": 22207,
         "file_name": "Intro_cut.mp4"
     },
     {
         "clip_number": 32,
-        "start_time_ms": 22692,
-        "end_time_ms": 23218,
+        "start_time_ms": 22712,
+        "end_time_ms": 23198,
         "file_name": "Conclusion_cut.mp4"
     }
 ]

dev.ipynb

@@ -208,7 +208,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 49,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -224,7 +224,7 @@
     "    {\n",
     "      \"character\": {\n",
     "        \"type\": \"avatar\",\n",
-    "        \"avatar_id\": \"f5db9448825249168273a87cd98f439d\",\n",
+    "        \"avatar_id\": \"42b4f9358cc9443a8610fda90bd3ad87\",\n",
     "        \"scale\": 1.0,\n",
     "        \"avatar_style\": \"normal\",\n",
     "        \"offset\": {\n",
@@ -245,14 +245,49 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 53,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "payload = {\n",
+    "    \"title\": \"Sample Talking Photo Video\",\n",
+    "    \"caption\": False,\n",
+    "    \"callback_id\": \"your_custom_id_123\",\n",
+    "    \"dimension\": {\n",
+    "        \"width\": 720,\n",
+    "        \"height\": 1280\n",
+    "    },\n",
+    "    \"video_inputs\": [\n",
+    "        {\n",
+    "            \"character\": {\n",
+    "                \"type\": \"talking_photo\",\n",
+    "                \"talking_photo_id\": 'cf557bdc461b487fb703b0c3138df325'\n",
+    "            },\n",
+    "            \"voice\": {\n",
+    "                \"type\": \"audio\",\n",
+    "                \"audio_url\": \"https://drive.google.com/file/d/1CF8TjWwPWiIPJX3D1bjpMopinHt4iw1D/view?usp=sharing\"\n",
+    "            },\n",
+    "            \"talking_style\": \"stable\",\n",
+    "            \"expression\": \"happy\",\n",
+    "            # \"background\": {\n",
+    "            #     \"type\": \"color\",\n",
+    "            #     \"value\": \"#FAFAFA\"\n",
+    "            # }\n",
+    "        }\n",
+    "    ]\n",
+    "}\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 54,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "{\"error\": null, \"data\": {\"video_id\": \"53cf5d6942454560971ba70926bdacc5\"}}\n"
+      "{\"error\": null, \"data\": {\"video_id\": \"d8194dfd76fa40c49bfc9a99c67ffbe3\"}}\n"
      ]
     }
    ],
@@ -277,7 +312,28 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 56,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'d8194dfd76fa40c49bfc9a99c67ffbe3'"
+      ]
+     },
+     "execution_count": 56,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "video_id = response.json()[\"data\"][\"video_id\"]\n",
+    "video_id"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 57,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -289,33 +345,33 @@
     "}\n",
     "\n",
     "\n",
-    "response = requests.get(url, headers=headers, params={\"video_id\": \"53cf5d6942454560971ba70926bdacc5\"}).json()"
+    "response = requests.get(url, headers=headers, params={\"video_id\": video_id}).json()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 58,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "text/plain": [
        "{'code': 100,\n",
-       " 'data': {'callback_id': 'your_custom_id_123',\n",
-       "  'caption_url': '',\n",
-       "  'created_at': 1736225867,\n",
-       "  'duration': 23.458,\n",
+       " 'data': {'callback_id': None,\n",
+       "  'caption_url': None,\n",
+       "  'created_at': 1736231063,\n",
+       "  'duration': None,\n",
        "  'error': None,\n",
-       "  'gif_url': 'https://resource2.heygen.ai/video/gifs/53cf5d6942454560971ba70926bdacc5.gif',\n",
-       "  'id': '53cf5d6942454560971ba70926bdacc5',\n",
-       "  'status': 'completed',\n",
-       "  'thumbnail_url': 'https://files2.heygen.ai/aws_pacific/avatar_tmp/a0cc99225ed34eb9811246368d94768a/53cf5d6942454560971ba70926bdacc5.jpeg?Expires=1736830809&Signature=kFj3rXlTETYCBXvi-8A37~-aVx1sZCBsDCzdU6W-mU0A45lsnGd1ao9v7QmkR5CqRkIgKNg9~zNKOmARVIHt9p37zpRN9uzcP3XUQ0wBS7C61pYZU-6KXxlfDbVt-aiM2BmvfhKZHA1W7p~fBgwYxGNt25LluA7hFLpczGswMVLEWb7z0JV84UfaZmU~5IqkCt-Fp-o69YlSZ26NvEjm~X2WpM2ILZ9bRpjrNkNNQ3LvUaO-~9XRhEEhe8ANeZgG1Jl7oh6dsQgLTC3TKwrFjsGMeXS2L3etIAja01iWdpEvbrW1oeWXSdPo8AW-0Zac-RJ4ZFymk8kJY6RiWDNxEQ__&Key-Pair-Id=K38HBHX5LX3X2H',\n",
-       "  'video_url': 'https://files2.heygen.ai/aws_pacific/avatar_tmp/a0cc99225ed34eb9811246368d94768a/53cf5d6942454560971ba70926bdacc5.mp4?Expires=1736830768&Signature=hub7aBI-57uvpQR-kL6eltYrFiacZxEOZKSKE60ntls~f2k87qacQNIiCbkXicKN3Q9CMyG0Io4vqMAvDapSrQXcaOfM6mOyBIV718LjCGXh9r35VVaMvbSl1onIIuoHeW2jYfeBVZ-mUgDHG~3ciHodXKX~VlebM595bHsg65LoBxYNVYEMJIHShxRdwE8xSmWahBUPubf~pegkCtJA1MBU0a7oN3jkey7EMlTDiBvRvXdES332mZkoNpFc7Ecfj6Lv-5k7T~ZBXq69bN79pnn4RVaw7vvsJiQnrKlrKYVe90HWb0MIVtjRkLX2cloWFhHpLh3picgprT0ixUhjsg__&Key-Pair-Id=K38HBHX5LX3X2H',\n",
+       "  'gif_url': None,\n",
+       "  'id': 'd8194dfd76fa40c49bfc9a99c67ffbe3',\n",
+       "  'status': 'processing',\n",
+       "  'thumbnail_url': None,\n",
+       "  'video_url': None,\n",
        "  'video_url_caption': None},\n",
        " 'message': 'Success'}"
       ]
      },
-     "execution_count": 21,
+     "execution_count": 58,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -414,7 +470,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 27,
+   "execution_count": 31,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -423,13 +479,81 @@
     "    clips_metadata = json.load(f)\n",
     "\n",
     "for item in clips_metadata:\n",
-    "    item[\"start_time_ms\"]=item[\"start_time_ms\"]+40\n",
-    "    item[\"end_time_ms\"]=item[\"end_time_ms\"]-40\n",
+    "    item[\"start_time_ms\"]=item[\"start_time_ms\"]+20\n",
+    "    item[\"end_time_ms\"]=item[\"end_time_ms\"]-20\n",
     "\n",
     "# update the file with new timestamps\n",
     "with open(\"clips_metadata.json\", \"w\") as f:\n",
     "    json.dump(clips_metadata, f, indent=4)"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 45,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "{'code': 100, 'data': {'talking_photo_id': 'cf557bdc461b487fb703b0c3138df325', 'talking_photo_url': 'https://files2.heygen.ai/prod/movio/url_upload/user_upload/a0cc99225ed34eb9811246368d94768a/f670829259c24a0197966e9f7c767ada.image/png?Expires=1736835408&Signature=EcHmA-5R6USvUeus0Ku~NCoCfg~T9NteKjfR59HEYOOurb7in9T4QB3Um8aeHBGLhIz1~5epJuvzIVnuB9H2Ab5uwV9B84KdXWK5rWhrCefRRn8JQxFOnvoCDtQFeHhuTs8957xbboJkFdLiv8OLli90lkmO0fzZecs7WB~WFFhQGqeAhCPjAvQt325ECSngNA2Z4MbI96YbOanE5GiRvUg56hczdb9XqhZktDkZKIDZGl3jk45BcmoaYh6N~sySnGN4kBe-Ny8exFzH0cjloX3gwkHnwsitmI2yQKmse7LKQFEt8WyQYxQ6z6q4YN1FpjKqMjK1T~~UGLepdGIDCA__&Key-Pair-Id=K38HBHX5LX3X2H'}, 'msg': None, 'message': None}\n"
+     ]
+    }
+   ],
+   "source": [
+    "api_key = \"YTBjYzk5MjI1ZWQzNGViOTgxMTI0NjM2OGQ5NDc2OGEtMTczMzk5MjA5MA==\"\n",
+    "\n",
+    "with open(\"man_portrait.png\", \"rb\") as f:\n",
+    "  resp = requests.post(\"https://upload.heygen.com/v1/talking_photo\", data=f, headers={\"Content-Type\": \"image/png\", \"x-api-key\": api_key})\n",
+    "  print(resp.json())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 52,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'cf557bdc461b487fb703b0c3138df325'"
+      ]
+     },
+     "execution_count": 52,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "response = resp.json()\n",
+    "response[\"data\"][\"talking_photo_id\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "url = \"https://api.heygen.com/v1/talking_photo.list\"\n",
+    "\n",
+    "headers = {\n",
+    "    \"accept\": \"application/json\",\n",
+    "    \"x-api-key\": api_key\n",
+    "}\n",
+    "\n",
+    "response = requests.get(url, headers=headers)\n",
+    "\n",
+    "print(response.text)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {

heygen.py

@@ -0,0 +1,173 @@
+import requests
+import json
+from moviepy.editor import VideoFileClip
+import os
+
+API_KEY = "YTBjYzk5MjI1ZWQzNGViOTgxMTI0NjM2OGQ5NDc2OGEtMTczMzk5MjA5MA=="
+AUDIO_URL = "https://drive.google.com/file/d/1CF8TjWwPWiIPJX3D1bjpMopinHt4iw1D/view?usp=sharing"
+
+
+# API ENDPOINTS
+#GET
+VIDEO_GENERATION_STATUS = "https://api.heygen.com/v1/video_status.get"
+LIST_TALKING_PHOTOS = "https://api.heygen.com/v1/talking_photo.list"
+
+#POST
+VIDEO_GENERATION = "https://api.heygen.com/v2/video/generate"
+UPLOAD_PHOTO = "https://upload.heygen.com/v1/talking_photo"
+
+
+def video_generation_payload(title, avatar_id, audio_url):
+    return {
+      "title": title,
+      "caption": False,
+    #   "callback_id": "callback_id",
+      "dimension": {
+        "width": 720,
+        "height": 1280
+      },
+      "video_inputs": [
+        {
+          "character": {
+            "type": "avatar",
+            "avatar_id": avatar_id,
+            "scale": 1.0,
+            "avatar_style": "normal",
+            "offset": {
+              "x": 0.0,
+              "y": 0.0
+            },
+            "matting": False
+          },
+          "voice": {
+            "type": "audio",
+            "audio_url": audio_url
+          },
+        }
+      ],
+    #   "callback_url": "callback_url"
+    }
+
+
+def talking_photo_payload(title, talking_photo_id, audio_url):
+    print(f"Creating video with title: {title}, talking_photo_id: {talking_photo_id}, audio_url: {audio_url}")
+    return {
+        "title": title,
+        "caption": False,
+        # "callback_id": "callback_id",
+        "dimension": {
+            "width": 720,
+            "height": 1280
+        },
+        "video_inputs": [
+            {
+                "character": {
+                    "type": "talking_photo",
+                    "talking_photo_id": talking_photo_id
+                },
+                "voice": {
+                    "type": "audio",
+                    "audio_url": audio_url
+                },
+                "talking_style": "stable",
+                "expression": "happy",
+            }
+        ],
+        # "callback_url": "callback_url"
+    }
+
+
+def upload_photo(file_path, type):
+    print(f"Uploading photo {file_path}")
+    with open(file_path, "rb") as f:
+        resp = requests.post(UPLOAD_PHOTO, data=f, headers={"Content-Type": type, "x-api-key": API_KEY}).json()
+        try:
+            talking_photo_id = resp["data"]["talking_photo_id"]
+            return talking_photo_id
+        except Exception as e:
+            print(f"Error uploading photo: {e}")
+            return resp
+
+def list_talking_photos():
+    print("Listing talking photos")
+    headers = {
+        "accept": "application/json",
+        "x-api-key": API_KEY
+    }
+    response = requests.get(LIST_TALKING_PHOTOS, headers=headers).json()
+    return response
+
+
+def download_video(video_id, video_url):
+    download_video = requests.get(video_url)
+    print(f"Downloading video {video_id}")
+    with open(f"{video_id}.mp4", "wb") as f:
+        f.write(download_video.content)
+
+
+def create_video(title, avatar_id, audio_url):
+    print(f"Creating video with title: {title}, avatar_id: {avatar_id}, audio_url: {audio_url}")
+    headers = {
+        "accept": "application/json",
+        "content-type": "application/json",
+        "x-api-key": API_KEY
+    }
+    payload = talking_photo_payload(title, avatar_id, audio_url)
+    response = requests.post(VIDEO_GENERATION, json=payload, headers=headers).json()
+
+    try:
+        video_id = response["data"]["video_id"]
+        print(f"Video ID: {video_id}")
+        return video_id
+    except Exception as e:
+        print(f"Error creating video: {e}")
+        return response
+
+
+def get_video_status(video_id):
+    headers = {
+        "accept": "application/json",
+        "x-api-key": API_KEY
+    }
+    response = requests.get(VIDEO_GENERATION_STATUS, headers=headers, params={"video_id": video_id}).json()
+
+    print(f"Video status: {response['data']['status']}")
+
+    return {
+        "id": response["data"]["id"],
+        "status": response["data"]["status"],
+        "video_url": response["data"]["video_url"],
+        "thumbnail_url": response["data"]["thumbnail_url"]
+    }
+
+
+def split_video(video_id):
+    print(f"Splitting video {video_id}.mp4")
+    with open("clips_metadata.json", "r") as f:
+        clips_metadata = json.load(f)
+
+    video = VideoFileClip(f"{video_id}.mp4")
+
+    clips = []
+    for clip_data in clips_metadata:
+        start_time = clip_data["start_time_ms"]
+        end_time = clip_data["end_time_ms"]
+
+        # convert milliseconds to seconds
+        start_time /= 1000
+        end_time /= 1000
+
+        try:
+            clip_name = clip_data["file_name"].split(".")[0]+".mp4"
+            clip = video.subclip(start_time, end_time)
+            clips.append(clip)
+        except Exception as e:
+            print(f"Error in clip {clip_name}: {e}")
+    
+    clips_dir = f"{video_id}_clips"
+    os.makedirs(clips_dir, exist_ok=True)
+    print(f"Saving clips to {clips_dir}")
+
+    for i, clip in enumerate(clips):
+        clip_path = os.path.join(clips_dir, clips_metadata[i]["file_name"].split(".")[0]+".mp4")
+        clip.write_videofile(clip_path, codec="libx264", audio_codec="aac")

utils.py

@@ -80,7 +80,7 @@ def cut_clip(clip, st, en):
 
 
 
-def create_video(number,st, en, output_file="output.mp4"):
+def create_video_old(number,st, en, output_file="output.mp4"):
     """
     Creates a video for the given number using pre-defined video clips.
     """
@@ -146,5 +146,35 @@ def create_advanced_video(number, trim_settings, avatar, output_file="output.mp4
     if video_clips:
         final_video = concatenate_videoclips(clips_to_combine)
         final_video.write_videofile(output_file, codec="libx264", audio_codec="aac")
+    else:
+        print("No clips to combine. Video creation failed.")
+
+
+def combine_video(number, video_id, output_file="output.mp4"):
+    path = f"{video_id}_clips"
+    intro_file = f"{path}/Intro_cut.mp4"
+    conclusion_file = f"{path}/Conclusion_cut.mp4"
+
+    clips_to_combine = [intro_file]
+
+    clips = generate_clip_sequence(number)
+    for clip in clips:
+        clips_to_combine.append(trimmed_clip_path(clip, avatar_path=path))
+
+    clips_to_combine.append(conclusion_file)
+
+    # Load and combine video clips
+    video_clips = []
+    for clip_path in clips_to_combine:
+        try:
+            clip = VideoFileClip(clip_path)
+            video_clips.append(clip)
+        except Exception as e:
+            print(f"Error loading clip {clip_path}: {e}")
+
+    # Concatenate the video clips
+    if video_clips:
+        final_video = concatenate_videoclips(video_clips)
+        final_video.write_videofile(output_file, codec="libx264", audio_codec="aac")
     else:
         print("No clips to combine. Video creation failed.")