07_food_vision_milestone_project_1 - how to get image name or id

[Onturk89]

Hello,

My steps are below:

import tensorflow_datasets as tfds
(train_data, test_data), ds_info = tfds.load(name="food101", # target dataset to get from TFDS
split=["train", "validation"],
shuffle_files=True,
as_supervised=True,
with_info=True)

We can get label name with the code below:
class_names = ds_info.features["label"].names

However I cannot find image name or id code related to train_data from tensorflow dataset.

I wish to build dataframe like below in order to see wrong predictions:

filepaths = []
for filepath in test_data.list_files("101_food_classes_10_percent/test//.jpg",
shuffle=False):
filepaths.append(filepath.numpy())

2. Create a dataframe out of current prediction data for analysis

import pandas as pd
pred_df = pd.DataFrame({"img_path": filepaths,
"y_true": y_labels,
"y_pred": pred_classes,
"pred_conf": pred_probs.max(axis=1), # get the maximum prediction probability value
"y_true_classname": [class_names[i] for i in y_labels],
"y_pred_classname": [class_names[i] for i in pred_classes]})
pred_df.head()

Any idea to get image name from food 101 dataset?

Thank you in advance.