# Combed DNA Image Recognition — Data

来源：https://www.kaggle.com/competitions/combed-dna-image-recognition/data

采集时间：2026-10-24T15:50:38Z

File descriptions
sampleSubmission.csv
- a sample submission file in the correct format
data.zip
: a zipped directory of images dataset organized as the following:
train: directory containing the training images divided into two classes '0' and '1'
test: directory containing the test images for prediction.
Note that the images base name should correspond to the images id in the submission file (ex: 23.tif  -->  23)
Images are encoded in 3-channels (BGR) 16 bits. For a simple visualization, we could use ImageJ (http://imagej.net) to adjust the contrast. A true signal image should look like the following one:
![True signal image in 8 bits](https://storage.googleapis.com/kaggle-media/competitions/inclass/4916/media/signal.png)
Data fields
id
- an anonymous id unique to a signal (the base name of the related image in the test directory)
label
- the label of the signal (0: noise, 1: true signal)