Roboflow + Make Sense
Connect a Roboflow model to Make Sense so it draws suggested boxes as you open each image, then upload the finished labels to Roboflow.
How to connect Make Sense to Roboflow
5 stepsLoad images into Make Sense
Open makesense.ai, drop in your images, choose object detection and start the project.
Connect your Roboflow model
Click Connect AI server and enter your model ID (project name and version number) and your private Roboflow API key.
Review the suggested boxes
Each image you open is sent to Roboflow and the predictions are drawn as boxes. Move, resize or delete them, and add any the model missed.
Export annotations
Click Actions, then Export Annotations, and choose VOC XML for boxes or COCO JSON for polygons.
Upload to Roboflow and retrain
Upload the images and exported labels to your Roboflow project, generate a new version and train an improved model.
What teams build with Make Sense
Label faster with your current model
Let a model you already trained draw first-pass boxes so labelers only fix mistakes.
Label without installing anything
Annotate in a browser tab and send the results to Roboflow when you are done.
Close the loop on a weak class
Label new images for the classes your model misses, upload them and retrain.
Questions
Object detection models served by the Roboflow Hosted API. Make Sense draws their predictions as bounding boxes.
VOC XML for bounding boxes, because each file carries the class names. Use COCO JSON if you drew polygons.
When the AI server is connected, each image you open is sent to Roboflow for inference. Without it, Make Sense works locally in the browser.
Why are companies choosing Roboflow?
The end-to-end platform for vision AI: create datasets, train models and deploy to production.
Flexible deployment modes
Move from a scalable cloud API to edge hardware like NVIDIA Jetson without managing complex dependencies.
Integrated workflow builder
Combine multiple AI models, custom business logic and third-party integrations into one production-ready pipeline in a low-code editor.
Model monitoring and alerts
Dashboards track confidence, inference time and request volume for your deployed models, and email alerts tell your team when confidence suddenly drops.
