Spaces:

---

Duplicated from  fffiloni/live-vision

awacke1

/

CameraLiveDocker

like 1

Sleeping

The human brain is very skilled at model making where your senses and feedback loops of action with memory cause us to build models in our brain as networks of neurons that activate in close timing and proximity.

I believe the capability of this AI to make repeated observations in a changing space that when accumulated and coalesced produces a memory of the situation in language. There may be emergent capabilities based on the combination of observations and prompts which can elicit constrained answers that determine whether the model is fit or if it lacks what we are looking for.

Prompt:

I have a video AI that can open your camera and as you point it at scenes it details the answer to a prompt question. For example "What is happening?" Develop a markdown outline with appropriate emojis to ask that will allow us to learn about what is in the scene and what it means to pain of things we want to undderstand, joy of when it provides verifiable facts and insights, and super power to be able to create a memory in the form of repeat output text files automatically named with a streamlit python UI that continually will list the text and image files captured that were the inference inpouts and outputs. The human brain is very skilled at model making where your senses and feedback loops of action with memory cause us to build models in our brain as networks of neurons that activate in close timing and proximity.

I believe the capability of this AI to make repeated observations in a changing space that when accumulated and coalesced produces a memory of the situation in language. There may be emergent capabilities based on the combination of observations and prompts which can elicit constrained answers that determine whether the model is fit or if it lacks what we are looking for.
Modify this program below to add these features and markdown outline. import express from "express";
import 'dotenv/config.js'
import { createServer } from "http";
import { Server } from "socket.io";
import { client } from "@gradio/client";
import { createRequire } from 'node:module'
const require = createRequire(import.meta.url);
global.EventSource = require('eventsource');

const app = express();
const httpServer = createServer(app);
app.use(express.static('public'));
const io = new Server(httpServer, { /* options */ });

io.on("connection", (socket) => {

console.log("new socket connection");

socket.on("ask_api", (client_data) => {
    console.log(client_data)
    console.log("trying to reach api");   
    asyncAPICall(client_data, socket)
});

});

async function test_servers(){
try{
const grapi_test = await client("https://gradio-hello-world.hf.space");
const apitest_result = await grapi_test.predict("/predict", [
"John",
]);
console.log(apitest_result);
}
catch(e){
console.log(e)
}

}

//test_servers()

async function asyncAPICall(data, socket) {

const grapi = await client("fffiloni/mndrm-call");
try{
const api_result = await grapi.predict("/infer", [
data[0], // blob in 'image' Image component
data[1], // string in 'Question' Textbox component
]);
console.log(api_result)
socket.emit("api_response", (api_result.data))
}
catch(e){
console.log(e)
socket.emit("api_error", ("ERROR ON API SIDE, SORRY..."))
}

}

httpServer.listen(7860);
console.log("App running on localhost:7860")

Response: