video-processing

Enhance Video Interactions with Apollo 7b Multiturn Actions

27 Apr 2025
Enhance Video Interactions with Apollo 7b Multiturn Actions

In today's digital landscape, engaging users through video content is more critical than ever. The Apollo 7b Multiturn service offers advanced Cognitive Actions designed to facilitate interactive multi-turn conversations with video content. This capability enhances user experience by allowing for detailed and contextual discussions, making it ideal for applications where understanding and engagement are paramount.

Imagine a scenario where a user is watching a tutorial video. Instead of passively consuming the content, they can ask questions about specific parts of the video, and receive real-time responses. This not only increases user engagement but also improves comprehension of the material. Such interactions can be applied in various fields, from education and training to customer support and interactive entertainment.

Prerequisites

To get started with the Apollo 7b Multiturn actions, you will need a Cognitive Actions API key and a basic understanding of making API calls.

Interactive Multi-Turn Video Chat

The Interactive Multi-Turn Video Chat action enables seamless, interactive conversations about video content. By utilizing an advanced model based on Apollo 7B, this action allows users to engage in detailed discussions, enhancing their overall experience with the video.

Input Requirements

To use this action, you must provide the following inputs:

  • Video: The URI of the input video file to be processed. For example, https://replicate.delivery/pbxt/M9kGHuJMeAKZs0eSbaEk6hCc7zqY4Tg94IxWwDpC5hRiuBPY/astro.mp4.
  • Messages: A JSON string that represents an array of messages, typically including user queries and assistant responses related to the video content. For instance, [{ "role": "user", "content": "Is there a dog in the video?" }].
  • Temperature: A numeric value controlling the randomness of the response generation, with a default of 0.4.
  • Max New Tokens: An integer that sets the cap on the number of new tokens generated in the response, defaulting to 256.
  • Top Probability: A numeric threshold for nucleus sampling, defaulting to 0.7.

Expected Output

The expected output for this action is a response based on the user’s queries about the video content. For example, if the user asks, "Is there a dog in the video?", the output could simply be "No".

Use Cases for this Specific Action

  • Education: Enhance learning experiences by allowing students to ask questions about instructional videos.
  • Customer Support: Enable customers to inquire about product features shown in demo videos, leading to better service.
  • Interactive Entertainment: Create engaging experiences in gaming or storytelling where users can interact with video narratives.
import requests
import json

# Replace with your actual Cognitive Actions API key and endpoint
# Ensure your environment securely handles the API key
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
# This endpoint URL is hypothetical and should be documented for users
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute"

action_id = "4cae4803-32b9-4144-bb34-26ccd14f2b33" # Action ID for: Interactive Multi-Turn Video Chat

# Construct the exact input payload based on the action's requirements
# This example uses the predefined example_input for this action:
payload = {
  "video": "https://replicate.delivery/pbxt/M9kGHuJMeAKZs0eSbaEk6hCc7zqY4Tg94IxWwDpC5hRiuBPY/astro.mp4",
  "messages": "[\n    {\"role\": \"user\", \"content\": \"Is there a dog in the video?\"}, \n    {\"role\": \"assistant\", \"content\": \"No\"},\n    {\"role\": \"user\", \"content\": \"Is there an astronaut in the video?\"}\n]",
  "temperature": 0.4,
  "maxNewTokens": 256,
  "topProbability": 0.7
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json",
    # Add any other required headers for the Cognitive Actions API
}

# Prepare the request body for the hypothetical execution endpoint
request_body = {
    "action_id": action_id,
    "inputs": payload
}

print(f"--- Calling Cognitive Action: {action.name or action_id} ---")
print(f"Endpoint: {COGNITIVE_ACTIONS_EXECUTE_URL}")
print(f"Action ID: {action_id}")
print("Payload being sent:")
print(json.dumps(request_body, indent=2))
print("------------------------------------------------")

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json=request_body
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully. Result:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body (non-JSON): {e.response.text}")
    print("------------------------------------------------")

Conclusion

The Apollo 7b Multiturn service offers a powerful tool for developers looking to enhance user engagement through interactive video conversations. By implementing the Interactive Multi-Turn Video Chat action, you can create applications that not only inform but also engage users in meaningful dialogue.

As you explore these capabilities, consider how they can be applied in your projects to create more immersive and interactive experiences. The next step is to integrate these actions into your applications and witness the transformation in user interaction with video content.

← Back to All Articles