Update app.py

app.py

@@ -1,4 +1,4 @@
-"""Visual Iterative Prompting Demo."""
+"""PIVOT Demo."""
 
 import gradio as gr
 import numpy as np
@@ -89,10 +89,10 @@ examples = [
 
 with gr.Blocks() as demo:
   gr.Markdown("""
-# Visual Iterative Prompting Demo
-The demo below showcases the Visual Iterative Prompting (VIP) algorithm.
+# PIVOT: Prompting with Iterative Visual Optimization
+The demo below showcases a version of the PIVOT algorithm, which uses iterative visual prompts to optimize and guide the reasoning of Vision-Langauge-Models (VLMs).
 Given an image and a description of an object or region, 
-VIP leverages a Vision-Language Model (VLM) to iteratively search for the point in the image that best corresponds to the description.
+PIVOT iteratively searches for the point in the image that best corresponds to the description.
 This is done through visual prompting, where instead of reasoning with text, the VLM reasons over images annotated with sampled points,
 in order to pick the best points.
 In each iteration, we take the points previously selected by the VLM, resample new points around the their mean, and repeat the process.
@@ -104,16 +104,16 @@ This demo uses GPT-4V, so it requires an OpenAI API key.
 To use the provided example images, you can right click on the image -> copy image, then click the clipboard icon in the Input Image box.
 
 Hyperparameters to set:
-* N Samples for Initialization - how many initial points are sampled for the first VIP iteration.
+* N Samples for Initialization - how many initial points are sampled for the first PIVOT iteration.
 * N Samples for Optimiazation - how many points are sampled for subsequent iterations.
 * N Iterations - how many optimization iterations to perform.
-* N Ensemble Recursions - how many ensembles for recursive VIP.
+* N Ensemble Recursions - how many ensembles for recursive PIVOT.
 
 Note that each iteration takes about ~10s, and each additional ensemble adds a multiple number of N Iterations.
 
-After VIP finishes, the image gallery below will visualize VIP results throughout all the iterations.
-There are two images for each iteration - the first one shows all the sampled points, and the second one shows which one VIP picked.
-The Info textbox will show the final selected pixel coordinate that VIP converged to.
+After PIVOT finishes, the image gallery below will visualize PIVOT results throughout all the iterations.
+There are two images for each iteration - the first one shows all the sampled points, and the second one shows which one PIVOT picked.
+The Info textbox will show the final selected pixel coordinate that PIVOT converged to.
 """.strip())
 
   gr.Markdown(