Hugging Face's logo

Spaces:
ludusc
/
latent-space-theories
Runtime error

App Files Community
latent-space-theories
File size: 10,611 Bytes 
058ceb7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
 
import streamlit as st
import streamlit.components.v1 as components

import dnnlib
import legacy

import pickle
import pandas as pd
import numpy as np
from pyvis.network import Network

import random
from sklearn.metrics.pairwise import cosine_similarity

from matplotlib.backends.backend_agg import RendererAgg

from backend.disentangle_concepts import *

_lock = RendererAgg.lock

HIGHTLIGHT_COLOR = '#e7bcc5'
st.set_page_config(layout='wide')


st.title('Comparison among concept vectors')
st.write('> **How do the concept vectors relate to each other?**')
st.write('> **What is their join impact on the image?**')
st.write("""Description to write""")

    
annotations_file = './data/vase_annotated_files/seeds0000-20000.pkl'
with open(annotations_file, 'rb') as f:
    annotations = pickle.load(f)

if 'image_id' not in st.session_state:
    st.session_state.image_id = 0
if 'concept_ids' not in st.session_state:
    st.session_state.concept_ids = ['Provenance ADRIA']
if 'space_id' not in st.session_state:
    st.session_state.space_id = 'Z'
if 'type_col' not in st.session_state:
    st.session_state.type_col = 'Provenance'

# def on_change_random_input():
#     st.session_state.image_id = st.session_state.image_id

# ----------------------------- INPUT ----------------------------------
st.header('Input')
input_col_1, input_col_2, input_col_3 = st.columns(3)
# --------------------------- INPUT column 1 ---------------------------
with input_col_1:
    with st.form('text_form'):
        
        # image_id = st.number_input('Image ID: ', format='%d', step=1)
        st.write('**Choose two options to disentangle**')
        type_col = st.selectbox('Concept category:', tuple(['Provenance', 'Shape Name', 'Fabric', 'Technique']))
        
        ann_df = pd.read_csv(f'./data/vase_annotated_files/sim_{type_col}_seeds0000-20000.csv')
        labels = list(ann_df.columns)
        labels.remove('ID')
        labels.remove('Unnamed: 0')
        
        concept_ids = st.multiselect('Concepts:', tuple(labels), default=[labels[2], labels[3]])

        st.write('**Choose a latent space to disentangle**')
        # chosen_text_id_input = st.empty()
        # concept_id = chosen_text_id_input.text_input('Concept:', value=st.session_state.concept_id)
        space_id = st.selectbox('Space:', tuple(['Z', 'W']))

        choose_text_button = st.form_submit_button('Choose the defined concept and space to disentangle')
        
        if choose_text_button:
            st.session_state.concept_ids = list(concept_ids)
            space_id = str(space_id)
            st.session_state.space_id = space_id
            st.session_state.type_col = type_col
        # st.write(image_id, st.session_state.image_id)

# ---------------------------- SET UP OUTPUT ------------------------------
epsilon_container = st.empty()
st.header('Output')
st.subheader('Concept vector')

# perform attack container
# header_col_1, header_col_2, header_col_3, header_col_4, header_col_5 = st.columns([1,1,1,1,1])
# output_col_1, output_col_2, output_col_3, output_col_4, output_col_5 = st.columns([1,1,1,1,1])
header_col_1, header_col_2 = st.columns([1,1])
output_col_1, output_col_2 = st.columns([1,1])

st.subheader('Derivations along the concept vector')

# prediction error container
error_container = st.empty()
smoothgrad_header_container = st.empty()

# smoothgrad container
smooth_head_1, smooth_head_2, smooth_head_3, smooth_head_4, smooth_head_5 = st.columns([1,1,1,1,1])
smoothgrad_col_1, smoothgrad_col_2, smoothgrad_col_3, smoothgrad_col_4, smoothgrad_col_5 = st.columns([1,1,1,1,1])

# ---------------------------- DISPLAY COL 1 ROW 1 ------------------------------
with output_col_1:
    vectors, nodes_in_common, performances = get_concepts_vectors(concept_ids, annotations, ann_df, latent_space=space_id)
    header_col_1.write(f'Concepts {", ".join(concept_ids)} - Latent space {space_id} - Relevant nodes in common: {nodes_in_common} - Performance of the concept vectors: {performances}')# - Nodes {",".join(list(imp_nodes))}')

    edges = []
    for i in range(len(concept_ids)):
        for j in range(len(concept_ids)):
            if i != j:
                print(f'Similarity between {concept_ids[i]} and {concept_ids[j]}')
                similarity = cosine_similarity(vectors[i,:].reshape(1, -1), vectors[j,:].reshape(1, -1))
                print(np.round(similarity[0][0], 3))
                edges.append((concept_ids[i], concept_ids[j], np.round(similarity[0][0], 3)))

    
    net = Network(height="750px", width="100%",)
    for e in edges:
        src = e[0]
        dst = e[1]
        w = e[2]

        net.add_node(src, src, title=src)
        net.add_node(dst, dst, title=dst)
        net.add_edge(src, dst, value=w, title=src + ' to ' + dst + ' similarity ' +str(w))
    
    # Generate network with specific layout settings
    net.repulsion(
                        node_distance=420,
                        central_gravity=0.33,
                        spring_length=110,
                        spring_strength=0.10,
                        damping=0.95
                       )

    # Save and read graph as HTML file (on Streamlit Sharing)
    try:
        path = '/tmp'
        net.save_graph(f'{path}/pyvis_graph.html')
        HtmlFile = open(f'{path}/pyvis_graph.html', 'r', encoding='utf-8')

    # Save and read graph as HTML file (locally)
    except:
        path = '/html_files'
        net.save_graph(f'{path}/pyvis_graph.html')
        HtmlFile = open(f'{path}/pyvis_graph.html', 'r', encoding='utf-8')

    # Load HTML file in HTML component for display on Streamlit page
    components.html(HtmlFile.read(), height=435)

with output_col_2:
    with open('data/CLIP_vecs_vases.pkl', 'rb') as f:
        vectors_CLIP = pickle.load(f)
        
    # st.write(f'Class ID {input_id} - {input_label}: {pred_prob*100:.3f}% confidence')
    #st.write('Concept vector', separation_vector)
    header_col_2.write(f'Concepts {", ".join(concept_ids)} - Latent space CLIP')# - Nodes {",".join(list(imp_nodes))}')

    edges_clip = []
    for c1 in concept_ids:
        for c2 in concept_ids:
            if c1 != c2:
                c1 = st.session_state.type_col + ' ' + c1
                c2 = st.session_state.type_col + ' ' + c2
                
                print(f'Similarity between {c1} and {c2}')
                similarity = cosine_similarity(vectors_CLIP[c1].reshape(1, -1), vectors_CLIP[c2].reshape(1, -1))
                print(np.round(similarity[0][0], 3))
                edges_clip.append((c1, c2, np.round(float(np.round(similarity[0][0], 3)), 3)))


    net_clip = Network(height="750px", width="100%",)
    for e in edges_clip:
        src = e[0]
        dst = e[1]
        w = e[2]

        net_clip.add_node(src, src, title=src)
        net_clip.add_node(dst, dst, title=dst)
        net_clip.add_edge(src, dst, value=w, title=src + ' to ' + dst + ' similarity ' +str(w))
    
    # Generate network with specific layout settings
    net_clip.repulsion(
                        node_distance=420,
                        central_gravity=0.33,
                        spring_length=110,
                        spring_strength=0.10,
                        damping=0.95
                       )

    # Save and read graph as HTML file (on Streamlit Sharing)
    try:
        path = '/tmp'
        net_clip.save_graph(f'{path}/pyvis_graph_clip.html')
        HtmlFile = open(f'{path}/pyvis_graph_clip.html', 'r', encoding='utf-8')

    # Save and read graph as HTML file (locally)
    except:
        path = '/html_files'
        net_clip.save_graph(f'{path}/pyvis_graph_clip.html')
        HtmlFile = open(f'{path}/pyvis_graph_clip.html', 'r', encoding='utf-8')

    # Load HTML file in HTML component for display on Streamlit page
    components.html(HtmlFile.read(), height=435)
    
# ----------------------------- INPUT column 2 & 3 ----------------------------        
with input_col_2:
   with st.form('image_form'):
        
        # image_id = st.number_input('Image ID: ', format='%d', step=1)
        st.write('**Choose or generate a random image to test the disentanglement**')
        chosen_image_id_input = st.empty()
        image_id = chosen_image_id_input.number_input('Image ID:', format='%d', step=1, value=st.session_state.image_id)
        
        choose_image_button = st.form_submit_button('Choose the defined image')
        random_id = st.form_submit_button('Generate a random image')

        if random_id:
            image_id = random.randint(0, 50000)
            st.session_state.image_id = image_id
            chosen_image_id_input.number_input('Image ID:', format='%d', step=1, value=st.session_state.image_id)
            
        if choose_image_button:
            image_id = int(image_id)
            st.session_state.image_id = int(image_id)
        # st.write(image_id, st.session_state.image_id)

with input_col_3:
    with st.form('Variate along the disentangled concepts'):
        st.write('**Set range of change**')
        chosen_epsilon_input = st.empty()
        epsilon = chosen_epsilon_input.number_input('Epsilon:', min_value=1, step=1)
        epsilon_button = st.form_submit_button('Choose the defined epsilon')

# # ---------------------------- DISPLAY COL 2 ROW 1 ------------------------------


with dnnlib.util.open_url('./data/vase_model_files/network-snapshot-010000.pkl') as f:
    model = legacy.load_network_pkl(f)['G_ema'].to('cpu') # type: ignore

if st.session_state.space_id == 'Z':
    original_image_vec = annotations['z_vectors'][st.session_state.image_id]
else:
    original_image_vec = annotations['w_vectors'][st.session_state.image_id]

img = generate_original_image(original_image_vec, model, latent_space=st.session_state.space_id)
# input_image = original_image_dict['image']
# input_label = original_image_dict['label']
# input_id = original_image_dict['id']

with smoothgrad_col_3:
    st.image(img)
    smooth_head_3.write(f'Base image')


images, lambdas = generate_joint_effect(model, original_image_vec, vectors, min_epsilon=-(int(epsilon)), max_epsilon=int(epsilon), latent_space=st.session_state.space_id)

with smoothgrad_col_1:
    st.image(images[0])
    smooth_head_1.write(f'Change of {np.round(lambdas[0], 2)}')

with smoothgrad_col_2:
    st.image(images[1])
    smooth_head_2.write(f'Change of {np.round(lambdas[1], 2)}')

with smoothgrad_col_4:
    st.image(images[3])
    smooth_head_4.write(f'Change of {np.round(lambdas[3], 2)}')

with smoothgrad_col_5:
    st.image(images[4])
    smooth_head_5.write(f'Change of {np.round(lambdas[4], 2)}')