Spaces:
Runtime error
Runtime error
File size: 24,196 Bytes
bfda67b d91bd79 3d08f5e d91bd79 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 |
---
title: YOLOV3 GradCAM
emoji: 🐢
colorFrom: pink
colorTo: green
sdk: gradio
sdk_version: 3.40.1
app_file: app.py
pinned: false
license: mit
---
# Gradio Object Detection App with GradCAM for YOLOv3 - ERAv1 Session 13
## Table of Contents
- [Introduction](#introduction)
- [Features](#features)
- [Model Performance](#model-performance)
- [Inference Samples](#inference-samples)
- [How to Use](#how-to-use)
- [Supported Classes](#supported-classes)
- [Link to the Model](#link-to-the-model)
- [Acknowledgements](#acknowledgements)
## Introduction
This Gradio app showcases an object detection model using YOLOv3 architecture. The model is trained with enhanced features like multi-resolution training and Mosaic Augmentation. Additionally, the app provides GradCAM outputs for better visualization of the model's predictions.
## Features
- **PytorchLightning Implementation**: The codebase has been refactored to use PytorchLightning for a more modular and scalable approach.
- **Multi-resolution Training**: Unlike traditional models that train on a fixed resolution, this model has been trained on multiple resolutions (416, 608, 896, 1280) for better generalization.
- **Mosaic Augmentation**: Implemented Mosaic Augmentation to enhance the training dataset, but only applied 75% of the time to maintain variety.
- **Precision Training**: The model is trained using float16 precision for faster convergence and reduced memory usage.
- **GradCAM Visualization**: Integrated GradCAM to provide a heatmap visualization of the regions in the image that the model focuses on during prediction.
## Model Performance
```
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ Validate metric ┃ DataLoader 0 ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ val_class_accuracy_epoch │ 81.89761352539062 │
│ val_loss │ 6.100630283355713 │
│ val_no_obj_accuracy_epoch │ 97.92534637451172 │
│ val_obj_accuracy_epoch │ 71.2684097290039 │
└───────────────────────────┴───────────────────────────┘
100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 619/619 [29:42<00:00, 2.88s/it]
MAP: 0.10860311985015869
```
## Inference Samples
## How to Use
1. Navigate to the Gradio app interface.
2. Upload a custom image or select from the provided samples.
3. Click on the "Predict" button.
4. View the object detection predictions along with the GradCAM heatmap.
## Supported Classes
## Model Architecture
```
----------------------------------------------------------------
Layer (type) Output Shape Param #
================================================================
Conv2d-1 [-1, 32, 416, 416] 864
BatchNorm2d-2 [-1, 32, 416, 416] 64
LeakyReLU-3 [-1, 32, 416, 416] 0
CNNBlock-4 [-1, 32, 416, 416] 0
Conv2d-5 [-1, 64, 208, 208] 18,432
BatchNorm2d-6 [-1, 64, 208, 208] 128
LeakyReLU-7 [-1, 64, 208, 208] 0
CNNBlock-8 [-1, 64, 208, 208] 0
Conv2d-9 [-1, 32, 208, 208] 2,048
BatchNorm2d-10 [-1, 32, 208, 208] 64
LeakyReLU-11 [-1, 32, 208, 208] 0
CNNBlock-12 [-1, 32, 208, 208] 0
Conv2d-13 [-1, 64, 208, 208] 18,432
BatchNorm2d-14 [-1, 64, 208, 208] 128
LeakyReLU-15 [-1, 64, 208, 208] 0
CNNBlock-16 [-1, 64, 208, 208] 0
ResidualBlock-17 [-1, 64, 208, 208] 0
Conv2d-18 [-1, 128, 104, 104] 73,728
BatchNorm2d-19 [-1, 128, 104, 104] 256
LeakyReLU-20 [-1, 128, 104, 104] 0
CNNBlock-21 [-1, 128, 104, 104] 0
Conv2d-22 [-1, 64, 104, 104] 8,192
BatchNorm2d-23 [-1, 64, 104, 104] 128
LeakyReLU-24 [-1, 64, 104, 104] 0
CNNBlock-25 [-1, 64, 104, 104] 0
Conv2d-26 [-1, 128, 104, 104] 73,728
BatchNorm2d-27 [-1, 128, 104, 104] 256
LeakyReLU-28 [-1, 128, 104, 104] 0
CNNBlock-29 [-1, 128, 104, 104] 0
Conv2d-30 [-1, 64, 104, 104] 8,192
BatchNorm2d-31 [-1, 64, 104, 104] 128
LeakyReLU-32 [-1, 64, 104, 104] 0
CNNBlock-33 [-1, 64, 104, 104] 0
Conv2d-34 [-1, 128, 104, 104] 73,728
BatchNorm2d-35 [-1, 128, 104, 104] 256
LeakyReLU-36 [-1, 128, 104, 104] 0
CNNBlock-37 [-1, 128, 104, 104] 0
ResidualBlock-38 [-1, 128, 104, 104] 0
Conv2d-39 [-1, 256, 52, 52] 294,912
BatchNorm2d-40 [-1, 256, 52, 52] 512
LeakyReLU-41 [-1, 256, 52, 52] 0
CNNBlock-42 [-1, 256, 52, 52] 0
Conv2d-43 [-1, 128, 52, 52] 32,768
BatchNorm2d-44 [-1, 128, 52, 52] 256
LeakyReLU-45 [-1, 128, 52, 52] 0
CNNBlock-46 [-1, 128, 52, 52] 0
Conv2d-47 [-1, 256, 52, 52] 294,912
BatchNorm2d-48 [-1, 256, 52, 52] 512
LeakyReLU-49 [-1, 256, 52, 52] 0
CNNBlock-50 [-1, 256, 52, 52] 0
Conv2d-51 [-1, 128, 52, 52] 32,768
BatchNorm2d-52 [-1, 128, 52, 52] 256
LeakyReLU-53 [-1, 128, 52, 52] 0
CNNBlock-54 [-1, 128, 52, 52] 0
Conv2d-55 [-1, 256, 52, 52] 294,912
BatchNorm2d-56 [-1, 256, 52, 52] 512
LeakyReLU-57 [-1, 256, 52, 52] 0
CNNBlock-58 [-1, 256, 52, 52] 0
Conv2d-59 [-1, 128, 52, 52] 32,768
BatchNorm2d-60 [-1, 128, 52, 52] 256
LeakyReLU-61 [-1, 128, 52, 52] 0
CNNBlock-62 [-1, 128, 52, 52] 0
Conv2d-63 [-1, 256, 52, 52] 294,912
BatchNorm2d-64 [-1, 256, 52, 52] 512
LeakyReLU-65 [-1, 256, 52, 52] 0
CNNBlock-66 [-1, 256, 52, 52] 0
Conv2d-67 [-1, 128, 52, 52] 32,768
BatchNorm2d-68 [-1, 128, 52, 52] 256
LeakyReLU-69 [-1, 128, 52, 52] 0
CNNBlock-70 [-1, 128, 52, 52] 0
Conv2d-71 [-1, 256, 52, 52] 294,912
BatchNorm2d-72 [-1, 256, 52, 52] 512
LeakyReLU-73 [-1, 256, 52, 52] 0
CNNBlock-74 [-1, 256, 52, 52] 0
Conv2d-75 [-1, 128, 52, 52] 32,768
BatchNorm2d-76 [-1, 128, 52, 52] 256
LeakyReLU-77 [-1, 128, 52, 52] 0
CNNBlock-78 [-1, 128, 52, 52] 0
Conv2d-79 [-1, 256, 52, 52] 294,912
BatchNorm2d-80 [-1, 256, 52, 52] 512
LeakyReLU-81 [-1, 256, 52, 52] 0
CNNBlock-82 [-1, 256, 52, 52] 0
Conv2d-83 [-1, 128, 52, 52] 32,768
BatchNorm2d-84 [-1, 128, 52, 52] 256
LeakyReLU-85 [-1, 128, 52, 52] 0
CNNBlock-86 [-1, 128, 52, 52] 0
Conv2d-87 [-1, 256, 52, 52] 294,912
BatchNorm2d-88 [-1, 256, 52, 52] 512
LeakyReLU-89 [-1, 256, 52, 52] 0
CNNBlock-90 [-1, 256, 52, 52] 0
Conv2d-91 [-1, 128, 52, 52] 32,768
BatchNorm2d-92 [-1, 128, 52, 52] 256
LeakyReLU-93 [-1, 128, 52, 52] 0
CNNBlock-94 [-1, 128, 52, 52] 0
Conv2d-95 [-1, 256, 52, 52] 294,912
BatchNorm2d-96 [-1, 256, 52, 52] 512
LeakyReLU-97 [-1, 256, 52, 52] 0
CNNBlock-98 [-1, 256, 52, 52] 0
Conv2d-99 [-1, 128, 52, 52] 32,768
BatchNorm2d-100 [-1, 128, 52, 52] 256
LeakyReLU-101 [-1, 128, 52, 52] 0
CNNBlock-102 [-1, 128, 52, 52] 0
Conv2d-103 [-1, 256, 52, 52] 294,912
BatchNorm2d-104 [-1, 256, 52, 52] 512
LeakyReLU-105 [-1, 256, 52, 52] 0
CNNBlock-106 [-1, 256, 52, 52] 0
ResidualBlock-107 [-1, 256, 52, 52] 0
Conv2d-108 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-109 [-1, 512, 26, 26] 1,024
LeakyReLU-110 [-1, 512, 26, 26] 0
CNNBlock-111 [-1, 512, 26, 26] 0
Conv2d-112 [-1, 256, 26, 26] 131,072
BatchNorm2d-113 [-1, 256, 26, 26] 512
LeakyReLU-114 [-1, 256, 26, 26] 0
CNNBlock-115 [-1, 256, 26, 26] 0
Conv2d-116 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-117 [-1, 512, 26, 26] 1,024
LeakyReLU-118 [-1, 512, 26, 26] 0
CNNBlock-119 [-1, 512, 26, 26] 0
Conv2d-120 [-1, 256, 26, 26] 131,072
BatchNorm2d-121 [-1, 256, 26, 26] 512
LeakyReLU-122 [-1, 256, 26, 26] 0
CNNBlock-123 [-1, 256, 26, 26] 0
Conv2d-124 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-125 [-1, 512, 26, 26] 1,024
LeakyReLU-126 [-1, 512, 26, 26] 0
CNNBlock-127 [-1, 512, 26, 26] 0
Conv2d-128 [-1, 256, 26, 26] 131,072
BatchNorm2d-129 [-1, 256, 26, 26] 512
LeakyReLU-130 [-1, 256, 26, 26] 0
CNNBlock-131 [-1, 256, 26, 26] 0
Conv2d-132 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-133 [-1, 512, 26, 26] 1,024
LeakyReLU-134 [-1, 512, 26, 26] 0
CNNBlock-135 [-1, 512, 26, 26] 0
Conv2d-136 [-1, 256, 26, 26] 131,072
BatchNorm2d-137 [-1, 256, 26, 26] 512
LeakyReLU-138 [-1, 256, 26, 26] 0
CNNBlock-139 [-1, 256, 26, 26] 0
Conv2d-140 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-141 [-1, 512, 26, 26] 1,024
LeakyReLU-142 [-1, 512, 26, 26] 0
CNNBlock-143 [-1, 512, 26, 26] 0
Conv2d-144 [-1, 256, 26, 26] 131,072
BatchNorm2d-145 [-1, 256, 26, 26] 512
LeakyReLU-146 [-1, 256, 26, 26] 0
CNNBlock-147 [-1, 256, 26, 26] 0
Conv2d-148 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-149 [-1, 512, 26, 26] 1,024
LeakyReLU-150 [-1, 512, 26, 26] 0
CNNBlock-151 [-1, 512, 26, 26] 0
Conv2d-152 [-1, 256, 26, 26] 131,072
BatchNorm2d-153 [-1, 256, 26, 26] 512
LeakyReLU-154 [-1, 256, 26, 26] 0
CNNBlock-155 [-1, 256, 26, 26] 0
Conv2d-156 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-157 [-1, 512, 26, 26] 1,024
LeakyReLU-158 [-1, 512, 26, 26] 0
CNNBlock-159 [-1, 512, 26, 26] 0
Conv2d-160 [-1, 256, 26, 26] 131,072
BatchNorm2d-161 [-1, 256, 26, 26] 512
LeakyReLU-162 [-1, 256, 26, 26] 0
CNNBlock-163 [-1, 256, 26, 26] 0
Conv2d-164 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-165 [-1, 512, 26, 26] 1,024
LeakyReLU-166 [-1, 512, 26, 26] 0
CNNBlock-167 [-1, 512, 26, 26] 0
Conv2d-168 [-1, 256, 26, 26] 131,072
BatchNorm2d-169 [-1, 256, 26, 26] 512
LeakyReLU-170 [-1, 256, 26, 26] 0
CNNBlock-171 [-1, 256, 26, 26] 0
Conv2d-172 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-173 [-1, 512, 26, 26] 1,024
LeakyReLU-174 [-1, 512, 26, 26] 0
CNNBlock-175 [-1, 512, 26, 26] 0
ResidualBlock-176 [-1, 512, 26, 26] 0
Conv2d-177 [-1, 1024, 13, 13] 4,718,592
BatchNorm2d-178 [-1, 1024, 13, 13] 2,048
LeakyReLU-179 [-1, 1024, 13, 13] 0
CNNBlock-180 [-1, 1024, 13, 13] 0
Conv2d-181 [-1, 512, 13, 13] 524,288
BatchNorm2d-182 [-1, 512, 13, 13] 1,024
LeakyReLU-183 [-1, 512, 13, 13] 0
CNNBlock-184 [-1, 512, 13, 13] 0
Conv2d-185 [-1, 1024, 13, 13] 4,718,592
BatchNorm2d-186 [-1, 1024, 13, 13] 2,048
LeakyReLU-187 [-1, 1024, 13, 13] 0
CNNBlock-188 [-1, 1024, 13, 13] 0
Conv2d-189 [-1, 512, 13, 13] 524,288
BatchNorm2d-190 [-1, 512, 13, 13] 1,024
LeakyReLU-191 [-1, 512, 13, 13] 0
CNNBlock-192 [-1, 512, 13, 13] 0
Conv2d-193 [-1, 1024, 13, 13] 4,718,592
BatchNorm2d-194 [-1, 1024, 13, 13] 2,048
LeakyReLU-195 [-1, 1024, 13, 13] 0
CNNBlock-196 [-1, 1024, 13, 13] 0
Conv2d-197 [-1, 512, 13, 13] 524,288
BatchNorm2d-198 [-1, 512, 13, 13] 1,024
LeakyReLU-199 [-1, 512, 13, 13] 0
CNNBlock-200 [-1, 512, 13, 13] 0
Conv2d-201 [-1, 1024, 13, 13] 4,718,592
BatchNorm2d-202 [-1, 1024, 13, 13] 2,048
LeakyReLU-203 [-1, 1024, 13, 13] 0
CNNBlock-204 [-1, 1024, 13, 13] 0
Conv2d-205 [-1, 512, 13, 13] 524,288
BatchNorm2d-206 [-1, 512, 13, 13] 1,024
LeakyReLU-207 [-1, 512, 13, 13] 0
CNNBlock-208 [-1, 512, 13, 13] 0
Conv2d-209 [-1, 1024, 13, 13] 4,718,592
BatchNorm2d-210 [-1, 1024, 13, 13] 2,048
LeakyReLU-211 [-1, 1024, 13, 13] 0
CNNBlock-212 [-1, 1024, 13, 13] 0
ResidualBlock-213 [-1, 1024, 13, 13] 0
Conv2d-214 [-1, 1024, 13, 13] 1,048,576
BatchNorm2d-215 [-1, 1024, 13, 13] 2,048
LeakyReLU-216 [-1, 1024, 13, 13] 0
CNNBlock-217 [-1, 1024, 13, 13] 0
Conv2d-218 [-1, 2048, 13, 13] 18,874,368
BatchNorm2d-219 [-1, 2048, 13, 13] 4,096
LeakyReLU-220 [-1, 2048, 13, 13] 0
CNNBlock-221 [-1, 2048, 13, 13] 0
Conv2d-222 [-1, 1024, 13, 13] 2,097,152
BatchNorm2d-223 [-1, 1024, 13, 13] 2,048
LeakyReLU-224 [-1, 1024, 13, 13] 0
CNNBlock-225 [-1, 1024, 13, 13] 0
Conv2d-226 [-1, 2048, 13, 13] 18,874,368
BatchNorm2d-227 [-1, 2048, 13, 13] 4,096
LeakyReLU-228 [-1, 2048, 13, 13] 0
CNNBlock-229 [-1, 2048, 13, 13] 0
ResidualBlock-230 [-1, 2048, 13, 13] 0
Conv2d-231 [-1, 1024, 13, 13] 2,097,152
BatchNorm2d-232 [-1, 1024, 13, 13] 2,048
LeakyReLU-233 [-1, 1024, 13, 13] 0
CNNBlock-234 [-1, 1024, 13, 13] 0
Conv2d-235 [-1, 2048, 13, 13] 18,874,368
BatchNorm2d-236 [-1, 2048, 13, 13] 4,096
LeakyReLU-237 [-1, 2048, 13, 13] 0
CNNBlock-238 [-1, 2048, 13, 13] 0
Conv2d-239 [-1, 75, 13, 13] 153,675
CNNBlock-240 [-1, 75, 13, 13] 0
ScalePrediction-241 [-1, 3, 13, 13, 25] 0
Conv2d-242 [-1, 256, 13, 13] 262,144
BatchNorm2d-243 [-1, 256, 13, 13] 512
LeakyReLU-244 [-1, 256, 13, 13] 0
CNNBlock-245 [-1, 256, 13, 13] 0
Upsample-246 [-1, 256, 26, 26] 0
Conv2d-247 [-1, 256, 26, 26] 196,608
BatchNorm2d-248 [-1, 256, 26, 26] 512
LeakyReLU-249 [-1, 256, 26, 26] 0
CNNBlock-250 [-1, 256, 26, 26] 0
Conv2d-251 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-252 [-1, 512, 26, 26] 1,024
LeakyReLU-253 [-1, 512, 26, 26] 0
CNNBlock-254 [-1, 512, 26, 26] 0
Conv2d-255 [-1, 256, 26, 26] 131,072
BatchNorm2d-256 [-1, 256, 26, 26] 512
LeakyReLU-257 [-1, 256, 26, 26] 0
CNNBlock-258 [-1, 256, 26, 26] 0
Conv2d-259 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-260 [-1, 512, 26, 26] 1,024
LeakyReLU-261 [-1, 512, 26, 26] 0
CNNBlock-262 [-1, 512, 26, 26] 0
ResidualBlock-263 [-1, 512, 26, 26] 0
Conv2d-264 [-1, 256, 26, 26] 131,072
BatchNorm2d-265 [-1, 256, 26, 26] 512
LeakyReLU-266 [-1, 256, 26, 26] 0
CNNBlock-267 [-1, 256, 26, 26] 0
Conv2d-268 [-1, 512, 26, 26] 1,179,648
BatchNorm2d-269 [-1, 512, 26, 26] 1,024
LeakyReLU-270 [-1, 512, 26, 26] 0
CNNBlock-271 [-1, 512, 26, 26] 0
Conv2d-272 [-1, 75, 26, 26] 38,475
CNNBlock-273 [-1, 75, 26, 26] 0
ScalePrediction-274 [-1, 3, 26, 26, 25] 0
Conv2d-275 [-1, 128, 26, 26] 32,768
BatchNorm2d-276 [-1, 128, 26, 26] 256
LeakyReLU-277 [-1, 128, 26, 26] 0
CNNBlock-278 [-1, 128, 26, 26] 0
Upsample-279 [-1, 128, 52, 52] 0
Conv2d-280 [-1, 128, 52, 52] 49,152
BatchNorm2d-281 [-1, 128, 52, 52] 256
LeakyReLU-282 [-1, 128, 52, 52] 0
CNNBlock-283 [-1, 128, 52, 52] 0
Conv2d-284 [-1, 256, 52, 52] 294,912
BatchNorm2d-285 [-1, 256, 52, 52] 512
LeakyReLU-286 [-1, 256, 52, 52] 0
CNNBlock-287 [-1, 256, 52, 52] 0
Conv2d-288 [-1, 128, 52, 52] 32,768
BatchNorm2d-289 [-1, 128, 52, 52] 256
LeakyReLU-290 [-1, 128, 52, 52] 0
CNNBlock-291 [-1, 128, 52, 52] 0
Conv2d-292 [-1, 256, 52, 52] 294,912
BatchNorm2d-293 [-1, 256, 52, 52] 512
LeakyReLU-294 [-1, 256, 52, 52] 0
CNNBlock-295 [-1, 256, 52, 52] 0
ResidualBlock-296 [-1, 256, 52, 52] 0
Conv2d-297 [-1, 128, 52, 52] 32,768
BatchNorm2d-298 [-1, 128, 52, 52] 256
LeakyReLU-299 [-1, 128, 52, 52] 0
CNNBlock-300 [-1, 128, 52, 52] 0
Conv2d-301 [-1, 256, 52, 52] 294,912
BatchNorm2d-302 [-1, 256, 52, 52] 512
LeakyReLU-303 [-1, 256, 52, 52] 0
CNNBlock-304 [-1, 256, 52, 52] 0
Conv2d-305 [-1, 75, 52, 52] 19,275
CNNBlock-306 [-1, 75, 52, 52] 0
ScalePrediction-307 [-1, 3, 52, 52, 25] 0
================================================================
Total params: 107,980,481
Trainable params: 107,980,481
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 1.98
Forward/backward pass size (MB): 1253.79
Params size (MB): 411.91
Estimated Total Size (MB): 1667.68
----------------------------------------------------------------
```
## Examples
App includes some examples images for testing
## Github
Training code may be found [here](https://github.com/Delve-ERAV1/S13)
## References
https://arxiv.org/abs/1804.02767 \
https://www.youtube.com/watch?v=Grir6TZbc1M \
https://github.com/jacobgil/pytorch-grad-cam
|