Jama automates waste contamination detection on collection vehicles with CVAT Online

Robotic surgery trades haptic feedback for precision. A surgeon at the console sees anatomy in extraordinary detail but cannot feel tissue consistency the way they would in open surgery. For trainees and early-career urologists, that absence of tactile information makes spatial orientation a genuine challenge: knowing exactly what structure lies behind the tissue you are cutting, and how far to go, demands experience that takes years to build.The Department of Urology at the University of Illinois Chicago (UIC) is building AI to give some of that orientation back. Led by Dr. Simone Crivellaro, one of the earliest adopters of the single-port (SP) robotic system worldwide, and Dr. Luca Morgantini, whose research focuses on AI-enabled decision support in robotic surgery, the team is training computer vision models to recognize anatomy in real time during robotic urological procedures. CVAT Online is the annotation infrastructure behind the dataset.What really separates a good robotic surgeon is never losing their orientation inside the patient and always knowing exactly what anatomy they're looking at. Real-time anatomy recognition is meant to give some of that orientation back.‍Luca MorgantiniMD Resident, Department of UrologyReal-time anatomy recognition in the robotic ORThe team's AI pipeline overlays anatomical labels directly onto the surgeon's console view during a robotic procedure. In transvesical enucleation of the prostate, the model recognizes three anatomical structures: the bladder neck, the prostatic adenoma, and the peripheral zone, tracking them frame by frame as the surgery progresses. The output appears through the console's TilePro feature, a second video channel built into the SP system, so the surgeon sees the AI's real-time assessment alongside the endoscopic view without interrupting the surgical workflow.The underlying model is trained on instance segmentation labels rather than bounding boxes, because the clinical value lies in knowing precisely where a structure begins and ends, not just its approximate location. YOLO is the primary architecture the team trains on, though U-Net has also been used depending on the task.Dr. Simone Crivellaro with the single port robot. Source. UIC was among the first programs worldwide to adopt the single-port SP robot when Intuitive, maker of the Da Vinci robotic system, introduced it in 2018, and it runs one of the few dedicated fellowships in minimally invasive urology. That combination of early clinical adoption and structured training pipeline is what makes the department both motivated and positioned to build this kind of surgical AI.Video embed: Provider Spotlight: Dr. Simone Crivellaro / Dr. Crivellaro - Uniqueness of daVinci SPThe challenge of annotating surgical anatomySurgical video is difficult to label. Every frame that goes into training must be extracted from a recorded procedure, de-identified, and annotated by someone who understands what they are looking at from a surgeon's perspective. Off-the-shelf computer vision annotators cannot interpret the structures inside a robotic procedure. The annotators have to have been at the console.“It matters a great deal that the people doing the annotation have actually been at the console and understand the anatomy first-hand,” says Dr. Morgantini. “The flip side is that we’re busy surgeons without much spare time for labeling, so that expertise is a real bottleneck.”The same expertise constraint shapes quality control. Surgical anatomy can be ambiguous in video: different annotators may segment the same structure differently depending on surgical phase, camera angle, or tissue state. The team handles this through a two-expert review protocol. Every annotation is confirmed by a second expert, and any genuine uncertainty is escalated to Dr. Crivellaro, who is internationally recognized in robotic urological surgery and trusted to resolve those conflicts definitively.The regulatory groundwork also cannot be skipped. IRB and institutional approval are required before any surgical video enters the annotation pipeline. Because every image is fully de-identified before upload, that approval is generally more straightforward to obtain. But it must come first.How UIC Urology uses CVAT OnlineThe annotation pipeline runs in three stages:Frame extraction and de-identification. Key frames are extracted from recorded SP robotic procedures and fully de-identified before any annotation work begins. The de-identified frames are uploaded directly into a CVAT Online project, where the whole team can begin working immediately.AI-assisted segmentation with SAM. CVAT Online's Segment Anything Model (SAM) integration runs a first-pass segmentation on each uploaded frame, generating anatomical outlines automatically. The expert annotator then reviews, adjusts, and confirms the AI-generated masks. Manual contouring of every frame by hand is replaced by review and correction, which is substantially faster given the complexity of surgical anatomy.Expert review and export. Each annotation is confirmed by a second expert. Verified labels are exported from CVAT Online in any required format. CVAT's multi-format export is a practical asset: the same labeled dataset can feed YOLO, U-Net, or other architectures without re-labeling, which matters as the team experiments across model types.The workflow turns the expert-time bottleneck into a manageable review task. Rather than asking a surgeon to draw every segmentation mask by hand, CVAT Online compresses their contribution to confirming what the model got right and correcting what it did not.Why choose CVAT Online?UIC arrived at CVAT Online through peer recommendation. When the team asked other medical AI researchers what annotation tool to use, CVAT came up consistently. "We got there the smart way," Dr. Morgantini notes, "as we asked other researchers what the most state-of-the-art option was, and they consistently pointed us to CVAT."The standout capability for surgical video annotation has been the SAM integration. Instance segmentation of complex anatomical structures requires tracing precise contours around irregular shapes in every frame. SAM reduces that to a review-and-confirm workflow, which is what makes the dataset feasible to build with a small team of expert annotators.Multi-user access has been equally important. The team spans faculty, residents, and collaborators across different locations, including international partners. CVAT Online allows everyone to work within the same project simultaneously. As Dr. Morgantini puts it: "Multiple teams across multiple sites collaborating on a single effort" is how research actually works now, and the platform reflects that.Cost was also a factor. CVAT Online came in less expensive than the commercial alternatives the team evaluated, which matters for a research program operating on grant funding.From annotated frames to the surgical consoleThe dataset has grown well beyond the initial pilot described in the team's publication, now exceeding 1,500 annotated surgical frames with annotation ongoing. A working anatomy-recognition model has been implemented through the TilePro feature on the SP console, overlaying real-time anatomical labels on the surgeon's view during procedures in research mode.The team has published a manuscript describing the pipeline and the results. The next planned study will test whether real-time AI anatomy guidance helps medical students recognize surgical structures faster than studying from textbooks alone, a step toward validating the tool in a structured educational setting before broader clinical adoption.The longer-term roadmap includes expansion across additional urological procedures, multi-institutional collaboration, and eventually intraoperative use in a clinical setting. The multi-user, cloud-based architecture of CVAT Online is already suited to that scale: annotators can contribute from anywhere, across time zones and institutions, to a single shared project.Building the annotation infrastructure for surgical AIUIC's work sits at the intersection of surgical training, clinical AI, and real-time decision support. The annotation pipeline built on CVAT Online is what allows a small team of expert surgeons to build training data at research scale, without sacrificing the domain-expertise standard that surgical AI demands.CVAT Online provides the multi-user access that makes distributed collaboration possible, the SAM integration that compresses manual labeling into focused review, and the flexible export that keeps the labeled data useful across architectures. As the program grows across procedures and institutions, the same infrastructure scales with it.

Getting meaningful coaching feedback in volleyball, or any sport, has always required access to a qualified coach. For the hundreds of millions of recreational and developing players who train without that access, technique errors go uncorrected and improvement stalls. Younalize was founded to change that. Built by developer Fabian van Schevikhoven and veteran volleyball coach Matt van Wezel in the Netherlands, the app uses computer vision and biomechanical analysis to deliver expert-level technique feedback to any player with a smartphone. CVAT Online sits at the heart of the data pipeline that makes it possible.Using computer vision to make sports coaching more accessibleThe idea began with co-founder Matt van Wezel, an experienced volleyball coach who has worked with national programs and teams in several countries. He initially approached co-founder Fabian van Schevikhoven with a marketplace concept: international players spend significant time traveling and waiting in airports, so perhaps they could be paid to review videos and give developing athletes technical feedback. The coaching method made sense. Matt often paused a video at the exact moment a player jumped, contacted the ball, or moved an elbow into the wrong position. A still frame gave the athlete a clear point of reference that was easier to understand than general feedback over a moving clip. The economics were harder. The athletes who most needed accessible coaching were not necessarily able to pay elite players for repeated reviews. That pushed the founders toward computer vision. Early experiments with pose estimation showed that body keypoints could be used to measure joint angles and movement phases. Ball detection added the context needed to understand when contact occurred. The concept evolved from a human review marketplace into an AI-assisted coach that could provide consistent feedback on demand. Bringing AI coaching to the volleyball courtYounalize is built around a four-step workflow. A player opens the app, selects a technique (jump serve, float serve, forearm pass, overhead pass, spike, or one of seven total techniques currently supported), records or uploads a short video clip, and within seconds receives annotated feedback overlaid directly on the footage.The analysis pipeline runs on a GPU server. When a video is uploaded, a pose detection model maps the player's body across every frame, tracking 133 keypoints from head to foot. In parallel, a custom object detection model locates the ball (whether yellow and blue, white, or beach-regulation red) even at the moment of contact, when motion blur is at its worst. The raw keypoint and ball position data is then passed to a rules engine, where a set of biomechanical conditions written by expert coaches determines which feedback to surface and what to say. For the jump serve alone, that rules engine contains more than 250 individual conditions, covering every phase of the movement from approach to follow-through.The result is coaching feedback that is both visual, with frame markers showing exactly where a joint is misaligned, and textual, surfacing a specific, actionable correction. The entire process, from video upload to annotated result, runs in seconds.The challenge of training on real-world sports footageThe obvious approach to AI sports coaching is pattern recognition: train a model on thousands of videos of correct technique and flag when a player deviates from the norm. Younalize evaluated this approach and ruled it out early, for a fundamental reason. The app's core users are recreational and developing players: athletes who are performing most techniques incorrectly, which is precisely why they are using the app. A pattern-matching model trained on professional footage has no stable reference point for a player who barely jumps, keeps their arm bent at contact, or has never learned to plant their feet before a serve.The team also tried training sequence-detection models to identify technique phases automatically. The variability across skill levels was too high. A dataset spanning elite club players and first-time beginners produced too much noise for any single model to segment reliably.Younalize's solution separates perception from coaching logic entirely. The AI handles what it does reliably: detecting and tracking bodies and objects in video. The coaching rules live in a separate layer, written by expert coaches and grounded in biomechanics rather than pattern matching. This architecture gives Younalize full control over the feedback the system surfaces, which an end-to-end ML approach cannot provide.How Younalize uses CVAT OnlineYounalize’s training data comes from footage that co-founder Matt van Wezel captures during coaching sessions, with player consent, across indoor courts, beach courts, and training facilities in the Netherlands, Norway, and India. The team has accumulated between 5,000 and 7,000 training videos spanning a wide range of playing levels, lighting conditions, and ball colors. None of this footage comes from the app's users: Younalize's privacy policy explicitly commits that uploaded videos are never used for model training.The annotation pipeline is managed entirely through CVAT Online, and works in three stages:Frame extraction and local annotation. Python scripts run locally to extract frames from raw video clips, filter out unusable footage (poor angle, excessive motion blur, or frames where the player is out of frame), and run Younalize's custom YOLO and SAM 2 models to generate bounding boxes and pose labels. The complete pre-annotated batch is then uploaded to CVAT Online for verification.Verification in CVAT Online. Fabian reviews each batch of tasks in CVAT Online, accepting accurate pre-annotations and correcting misses, most commonly ball detections lost to motion blur or occlusion.Export, and retraining. Verified annotations are exported and combined with the corresponding frames to produce a new training dataset, which runs overnight on a dedicated GPU server. The next round of pre-annotations reflects the improved model.The loop runs continuously alongside Fabian and Matt’s other responsibilities. A batch of four or five video clips typically takes less than half an hour to verify. By the following morning, a new model version is ready.Why choose CVAT Online?Younalize initially ran CVAT Community, the open-source edition, on Fabian's laptop via Docker. The setup worked for solo annotation, but collaboration with Matt van Wezel proved impossible: copying project directories between machines failed repeatedly, and the two founders ended up annotating independently without a shared dataset. Moving to CVAT Online resolved the problem immediately: both founders could access the same projects, tasks, and job queues from wherever they were working.CVAT Online's API gives Younalize the flexibility to integrate their own custom models directly into the annotation pipeline as their workflow evolves. The team has used it to call their YOLO and SAM 2 models before frames reach the manual review stage, an approach that keeps human effort focused on the genuinely hard cases: motion-blurred ball detections, occluded keypoints, unusual playing conditions. The combined result is measurable: dataset preparation that previously took weeks now takes days. For a two-person team without dedicated annotation staff, that compression is the difference between shipping model improvements continuously and batching them into infrequent labeling sprints.1,300 users, 90 countries, no marketing budgetYounalize reached 1,300 active users across 90 countries without a paid marketing campaign. One national volleyball federation has already purchased 400 licenses for their pre-youth national training program. Another federation embeds Younalize in its annual coach development course. Each year, 100 to 150 trainers learn to run technique analyses on their players and design training plans from the feedback the app returns.The app is also being used as a remote coaching tool. A Dutch head coach working with an African national beach volleyball team uses Younalize to monitor player technique during the months he spends back in the Netherlands. Players upload their own clips; the coach reviews AI-generated feedback remotely and builds individualized training plans from the results.These use cases reflect the wider opportunity Younalize is pursuing. The International Volleyball Federation (FIVB) aims to grow the global volleyball movement from an estimated 800 million people today to 1.6 billion by 2032, according to its Strategic Vision 2032. The large majority of that growth will come from markets where professional coaching is unavailable or unaffordable. Younalize is in active conversations with the federation about embedding the app in their Volleyball Empowerment program, which places elite coaches in underserved markets to develop the sport from the ground up.The annotation workflow built on CVAT Online is already supporting this expansion. Field hockey is in active development, requiring new detection models for both the ball and the hockey stick. No reliable pre-trained models existed for either, so Younalize built them from scratch using the same pipeline. A strength and conditioning module is in progress for athletes in national training programs. Each new sport or technique follows the same path: collect footage, annotate in CVAT Online, train, iterate.Making expert coaching accessible at scaleYounalize is building the infrastructure to make expert sports coaching as accessible as a smartphone. The annotation pipeline that powers the AI, running on CVAT Online and driven by custom pre-annotation models managed by two founders, is what allows the product to improve continuously without a dedicated data team.CVAT Online provides the cloud access that makes collaboration possible, the API flexibility that lets Younalize's own models do the pre-annotation work, and the annotation environment where human judgment is applied only where it is genuinely needed. As Younalize scales across new techniques, new sports, and new markets, the same workflow scales with it.