add minimal dataset card with link to paper

#5
by davanstrien HF staff - opened
Files changed (1) hide show
  1. README.md +158 -1
README.md CHANGED
@@ -122,7 +122,7 @@ dataset_info:
122
  '1': 'Yes'
123
  splits:
124
  - name: test
125
- num_bytes: 15119280526.0
126
  num_examples: 53304
127
  - name: validation
128
  num_bytes: 5013154770.625
@@ -138,4 +138,161 @@ configs:
138
  path: data/test-*
139
  - split: validation
140
  path: data/validation-*
 
 
 
 
 
141
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
122
  '1': 'Yes'
123
  splits:
124
  - name: test
125
+ num_bytes: 15119280526
126
  num_examples: 53304
127
  - name: validation
128
  num_bytes: 5013154770.625
 
138
  path: data/test-*
139
  - split: validation
140
  path: data/validation-*
141
+ task_categories:
142
+ - image-classification
143
+ pretty_name: Wake Vision
144
+ size_categories:
145
+ - 1M<n<10M
146
  ---
147
+
148
+ # Dataset Card for Wake Vision
149
+
150
+
151
+ ## Dataset Details
152
+
153
+ ### Dataset Description
154
+
155
+ <!-- Provide a longer summary of what this dataset is. -->
156
+
157
+ Paper abstract:
158
+
159
+ >Abstract. Machine learning applications on extremely low-power de-
160
+ vices, commonly referred to as tiny machine learning (TinyML), promises
161
+ a smarter and more connected world. However, the advancement of cur-
162
+ rent TinyML research is hindered by the limited size and quality of per-
163
+ tinent datasets. To address this challenge, we introduce Wake Vision, a
164
+ large-scale, diverse dataset tailored for person detection—the canonical
165
+ task for TinyML visual sensing. Wake Vision comprises over 6 million
166
+ images, which is a hundredfold increase compared to the previous stan-
167
+ dard, and has undergone thorough quality filtering. Using Wake Vision
168
+ for training results in a 2.41% increase in accuracy compared to the estab-
169
+ lished benchmark. Alongside the dataset, we provide a collection of five
170
+ detailed benchmark sets that assess model performance on specific seg-
171
+ ments of the test data, such as varying lighting conditions, distances from
172
+ the camera, and demographic characteristics of subjects. These novel
173
+ fine-grained benchmarks facilitate the evaluation of model quality in chal-
174
+ lenging real-world scenarios that are often ignored when focusing solely
175
+ on overall accuracy. Through an evaluation of a MobileNetV2 TinyML
176
+ model on the benchmarks, we show that the input resolution plays a
177
+ more crucial role than the model width in detecting distant subjects and
178
+ that the impact of quantization on model robustness is minimal, thanks
179
+ to the dataset quality. These findings underscore the importance of a de-
180
+ tailed evaluation to identify essential factors for model development. The
181
+ dataset, benchmark suite, code, and models are publicly available under
182
+ the CC-BY 4.0 license, enabling their use for commercial use cases
183
+
184
+ - **Curated by:** [More Information Needed]
185
+ - **Funded by [optional]:** [More Information Needed]
186
+ - **Shared by [optional]:** [More Information Needed]
187
+ - **Language(s) (NLP):** [More Information Needed]
188
+ - **License:** [More Information Needed]
189
+
190
+ ### Dataset Sources [optional]
191
+
192
+ <!-- Provide the basic links for the dataset. -->
193
+
194
+ - **Repository:** [More Information Needed]
195
+ - **Paper [optional]:** [More Information Needed]
196
+ - **Demo [optional]:** [More Information Needed]
197
+
198
+ ## Uses
199
+
200
+ <!-- Address questions around how the dataset is intended to be used. -->
201
+
202
+ ### Direct Use
203
+
204
+ <!-- This section describes suitable use cases for the dataset. -->
205
+
206
+ [More Information Needed]
207
+
208
+ ### Out-of-Scope Use
209
+
210
+ <!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
211
+
212
+ [More Information Needed]
213
+
214
+ ## Dataset Structure
215
+
216
+ <!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
217
+
218
+ [More Information Needed]
219
+
220
+ ## Dataset Creation
221
+
222
+ ### Curation Rationale
223
+
224
+ <!-- Motivation for the creation of this dataset. -->
225
+
226
+ [More Information Needed]
227
+
228
+ ### Source Data
229
+
230
+ <!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
231
+
232
+ #### Data Collection and Processing
233
+
234
+ <!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
235
+
236
+ [More Information Needed]
237
+
238
+ #### Who are the source data producers?
239
+
240
+ <!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
241
+
242
+ [More Information Needed]
243
+
244
+ ### Annotations [optional]
245
+
246
+ <!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
247
+
248
+ #### Annotation process
249
+
250
+ <!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
251
+
252
+ [More Information Needed]
253
+
254
+ #### Who are the annotators?
255
+
256
+ <!-- This section describes the people or systems who created the annotations. -->
257
+
258
+ [More Information Needed]
259
+
260
+ #### Personal and Sensitive Information
261
+
262
+ <!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
263
+
264
+ [More Information Needed]
265
+
266
+ ## Bias, Risks, and Limitations
267
+
268
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
269
+
270
+ [More Information Needed]
271
+
272
+ ### Recommendations
273
+
274
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
275
+
276
+
277
+ ## Citation
278
+
279
+ <!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
280
+
281
+ **BibTeX:**
282
+
283
+ ```bibtex
284
+ @misc{banbury2024wake,
285
+ title={Wake Vision: A Large-scale, Diverse Dataset and Benchmark Suite for TinyML Person Detection},
286
+ author={Colby Banbury and Emil Njor and Matthew Stewart and Pete Warden and Manjunath Kudlur and Nat Jeffries and Xenofon Fafoutis and Vijay Janapa Reddi},
287
+ year={2024},
288
+ eprint={2405.00892},
289
+ archivePrefix={arXiv},
290
+ primaryClass={cs.CV}
291
+ }
292
+ ```
293
+
294
+ [More Information Needed]
295
+
296
+ ## Dataset Card Contact
297
+
298
+ [More Information Needed]