# Using a Deep Learning Tool ## Deep Learning Studio * Include the new images in the test set * Add new images to the dataset from the `Dataset and annotations` tab. * In the `Validation and results` tab, refresh the results ![image](assets/images/easydeeplearning/easyclassify_6.webp) * Use the `Inference tests` tab to classify the new images and display the detailed results for these images. * You can also export the results to CSV from this tab. ## API * The API to use a **Deep Learning** tool on a new image or on a set of images is different for each tool: * **EasyClassify** > [Classifying New Images](../easyclassify/index.html.md#easyclassify-inference) * **EasySegment Unsupervised** > [Applying the Tool to New Images](../easysegment/easysegment_unsupervised.html.md#easysegment-unsupervised-inference) * **EasySegment Supervised** > [Applying the Tool to New Images](../easysegment/easysegment_supervised.html.md#easysegment-supervised-inference) * **EasyLocate** > [Applying the Tool to New Images](../easylocate/index.html.md#easylocate-inference) * You can apply each tool to individual images or to sets of images. * With a set of images and with GPU processing enabled, the tool processes [`EDeepLearningTool::GetBatchSize`](../../../doxygen_cpp/classeVision_1_1EasyDeepLearning_1_1EDeepLearningTool.html#a08067d6a1df8bc6767e2072c4efde09a) [.NET](../../../doxygen_cs/classeVision_1_1EasyDeepLearning_1_1EDeepLearningTool.html#a5d6d6c5c316f2171cc101e09d6008f85) [Python](../../../doxygen_python/classeVision_1_1EasyDeepLearning_1_1EDeepLearningTool.html#a6faea0ff36adc9568c36fc73268d58f1) at the same time. * **EasySegment** can automatically split individual images in sub-images according to the patch size, the scaling and the sampling density parameters. These sub-images are processed by batch. * Use the methods [`EDeepLearningTool::InitializeInference`](../../../doxygen_cpp/classeVision_1_1EasyDeepLearning_1_1EDeepLearningTool.html#a5a70b323809870d9a2512d46442718b0) [.NET](../../../doxygen_cs/classeVision_1_1EasyDeepLearning_1_1EDeepLearningTool.html#a0c889e8f70fedd86e77496edfd55a3ad) [Python](../../../doxygen_python/classeVision_1_1EasyDeepLearning_1_1EDeepLearningTool.html#a841de09ff474b83dd9774f49cdf7d32c) with an image or a set of images to initialize the inference engine after changing the batch size, the execution settings, or loading a model. Otherwise, the initialization will be performed on the first prediction. The initialization can take a significant amount of time.