Stereo Commercial Kitchen Line Cooking Data

It is first-person video of a person doing commercial kitchen work, recorded from their own point of view with two calibrated cameras and with hand pose and sensor streams alongside. The tasks are the rows the site’s catalog lists in Restaurant & Commercial Kitchen. Stereo describes the capture, and MCAP describes the container. They are separate choices.

The site does not state that stereo is already running on these rows. The catalog lists a head-mounted phone or a multiview rig as the capture setup, and a stereo pair is a capture choice agreed in the skill spec and confirmed on a Spec Pilot. The data is human capture by paid, consented contributors, not robot or teleoperation data. Firsthand is not affiliated with Foxglove or the MCAP maintainers.

Two, both in Restaurant & Commercial Kitchen under Food & Beverage. The catalog lists examples for the environment such as a sit-down restaurant kitchen, a café, a coffee shop, an artisan bakery and a patisserie. "Culinary Chef Work / Knife Work" is described as breaking down produce and proteins with precise knife cuts and cooking dishes on a professional line, and is rated Hard. The vegetable row, listed as chopping and portioning vegetables using a chef’s knife and cutting board, is described as washing, trimming and chopping vegetables into even pieces and portioning them into containers, and is rated Medium.

Both list a head-mounted phone or a multiview rig as the capture setup, and both are Active collection, exactly as the catalog states. The site says nothing further about what that means in volume, hours or delivered stereo footage, so this page claims none, and what applies to a given order is confirmed when you request the quote. The same environment also lists a "Cooking / Line Cooking" row, described as front and back of house restaurant operations including prep, plating and service, rated Medium and Active collection; this page does not rely on it. This page names only the tools and steps that appear in the two rows: a chef’s knife, a cutting board, produce, proteins, vegetables, containers, a professional line and knife cuts. The first table below reproduces the rows as the catalog states them.

The egocentric cutting and food prep page covers both of these rows as first-person capture. Its capture-stack section describes a head-mounted stereo rig as the instrumented tier, and it has no MCAP treatment. The egocentric cooking video page is wider, covering cooking video in general, with the same stereo rig mention and no MCAP treatment. The food-service egocentric page covers food-service settings and has no stereo or MCAP treatment.

This page is the stereo and MCAP treatment of the knife work and chopping rows, with the line cooking context from the culinary row. It does not repeat the baking, sandwich, sushi or cleaning rows those pages link to, and the stereo commercial laundry and linen page covers a different environment. The comparison table below sets the pages side by side.

Knife work is about where the blade, the ingredient and the guiding hand are relative to one another: how far the blade is from the fingertips, where the cut lands on the piece, whether the pieces are even. An ingredient also changes shape with every cut. A stereo pair recovers depth from the calibrated frame pair rather than from a standalone depth sensor, so the blade, the ingredient and the hands share a frame.

The site gives no depth accuracy figure and no label for an even piece, a cut line or a finished dish, so whether those are resolvable in depth is Not stated. It does not state how a stereo pair behaves on a blade, a steaming pan or a wet board either, and it says depth degrades on wet and reflective surfaces. Which kitchens a given program uses is agreed in the spec. Capture at your own facility is listed as not offered.

The site describes two stereo setups and does not say which one a kitchen order uses. The hard-environments stack is a global-shutter 4K/60 head camera with a 120° lens, two 1080p/60 wrist cameras, a depth sensor at 848×480 and 30 fps, and 200 Hz 6-DoF IMUs. The off-the-shelf stereo head rig, dataset 06, is dual fisheye cameras at 1080p per eye, a 62 mm baseline, per-unit calibration, a 9-axis IMU and hand-tracking output.

Which rig, resolution and streams a kitchen order uses is Not stated on the site, and the third table below sets what is stated beside what is not. Intrinsics and extrinsics are re-solved at the start and end of every session, and streams are hardware-triggered onto one clock with a 2.0 ms cross-stream ceiling. The catalog states that a multiview rig adds synchronized third-person angles when the spec asks for them, which a spec can use for a line the head camera sees only in part, and wrist cameras give a closer view of the hands.

MCAP is an open source container file format for multimodal log data, supporting multiple channels of timestamped pre-serialized data. It is serialization-agnostic, so messages can be Protobuf, DDS (CDR), ROS, JSON and more. Its design is row-oriented and append-only, it stores schemas alongside the data, and it offers an optional index plus optional LZ4 or Zstandard chunk compression.

Reader and writer libraries exist for C++, Go, Python, Rust, Swift and TypeScript, and the mcap CLI can inspect, merge and split files. The foxglove/mcap repository is MIT-licensed; that is the MCAP project’s licence, not the licence of a Firsthand delivery. The MCAP hub page and the glossary entry cover the format in full.

Firsthand offers MCAP delivery, and dataset 06 is stated as shipping as MCAP. For kitchen work the site states hand pose, a stereo-derived depth stream, IMU, contact events, action segments and per-frame instance masks elsewhere. What it does not state is which of those are channels in an MCAP file, how they are named, what encoding they use or at what rate.

No topic name, message schema or rate is stated on this page, and none should be assumed. Name the streams that matter in the skill spec and confirm them on a Spec Pilot batch. The fourth table below sets each stream beside what the site states about it.

Changing shape is the first risk. A cut ingredient is a different object from the whole one, so the thing being tracked changes between frames. The site ships per-frame instance masks rather than propagated ones for deformable objects, and it lists no cut-size, evenness, knife-angle, doneness or plating label, so what counts as a finished cut or a finished dish is written into the spec up front.

Hands are the second. One hand holds and guides the ingredient while the other holds the knife, and fingers curl behind the piece. The visibility flag on each joint is the answer the site gives, and contact events mark when a hand meets an item. A spec can also ask for wrist views, so a grip the head camera loses is still recorded.

Calibration is the third, because stereo depth is only as good as the rectification behind it: calibration is measured at the start and end of every session under a 0.28 px RMS gate. An episode is also rejected when the manipulated object is out of frame for more than 15 percent of it or an undocumented bystander appears, and a reject log ships with every batch. Force and tactile readings are not among the listed streams.

For every delivered episode the site states that faces, plates and screens are blurred irreversibly, and that footage that cannot be de-identified is discarded. Who may be in frame, and how a kitchen’s own staff, menus or labels are handled, is agreed in the spec.

The site states no cuisine, recipe, throughput, accuracy, food-safety or certification claim for these rows, so this page describes none.

Off-the-shelf covers what is already in stock: the stereo category lists the stereo datasets and hand-tracking video, and only dataset 06 is stated as MCAP. This page does not claim off-the-shelf stereo coverage of the kitchen rows. Custom collection covers the rest: custom collection at scale through a network of thousands of capture providers, for a consistent, recurring data flow to labs.

Where coverage is deep, the median time from signed spec to first validated batch is a median of 48 hours; where it is thin, turnaround runs three to four weeks. The site does not say which speed applies to a stereo kitchen order, so ask when you request the quote. Collections are billed per validated hour.

Describe the stations and their order, the ingredients, knives, boards and containers and their start state, what counts as a finished step, and the variation your pipeline is missing. Then name the streams you want in the file, the topic naming and message encoding your reader expects, whether calibration ships inside the file or beside it, and whether MCAP is the only delivery or sits beside RLDS, LeRobot or WebDataset.

Most programs begin with a Spec Pilot, a small first batch that proves the spec and accept criteria on real footage before you scale. The fifth table below shows what each spec element decides.