app.py

from huggingface_hub import from_pretrained_keras
import numpy as np
import json
import gradio as gr
import tensorflow_text
import tensorflow_addons
import tensorflow as tf
import matplotlib.pyplot as plt
import matplotlib.image as mpimg

# load config
with open("image_paths.json", 'r') as f:
    image_paths = json.load(f)
image_embeddings = np.load("image_embeddings.npy")
text_encoder = from_pretrained_keras("keras-io/dual-encoder-image-search")

def find_matches(image_paths, image_embeddings, queries, k=9, normalize=True):
    # Get the embedding for the query.
    query_embedding = text_encoder(tf.convert_to_tensor(queries))
    # Normalize the query and the image embeddings.
    if normalize:
        image_embeddings = tf.math.l2_normalize(image_embeddings, axis=1)
        query_embedding = tf.math.l2_normalize(query_embedding, axis=1)
    # Compute the dot product between the query and the image embeddings.
    dot_similarity = tf.matmul(query_embedding, image_embeddings, transpose_b=True)
    # Retrieve top k indices.
    results = tf.math.top_k(dot_similarity, k).indices.numpy()
    # Return matching image paths.
    return [[image_paths[idx] for idx in indices] for indices in results]


def inference(query):
    matches = find_matches(image_paths, image_embeddings, [query], normalize=True)[0]
    plt.figure(figsize=(20, 20))
    for i in range(9):
        ax = plt.subplot(3, 3, i + 1)
        plt.imshow(mpimg.imread(matches[i]))
        plt.axis("off") 
    plt.savefig("img.png")
    return "img.png"

examples= ['a family standing next to the ocean on a sandy beach with a surf board',
           'a group of people sitting in an audience with pen and paper',
           'a couple of cows that are in some grass'         
          ]

gr.Interface(
    fn=inference,
    title="Natural language image search with a Dual Encoder",
    description = "Implementation of a dual encoder model for retrieving images that match natural language queries (Note: for demo purposes, only 1k images were used as search space)",
    inputs="text",
    examples=examples,
    outputs="image",
    cache_examples=False,
    article = "Author: <a href=\"https://huggingface.co/vumichien\">Vu Minh Chien</a>. Based on the keras example from <a href=\"https://keras.io/examples/nlp/nl_image_search/\">Khalid Salama</a>",
).launch(debug=True, enable_queue=True)