Stereo Knitting and Crochet Data

It is real-world, first-person video of a person doing knitting or crochet work, recorded with two calibrated cameras so that depth can be recovered, together with 3D hand pose and the other streams the site already lists. The catalog describes the work in its own words: hand and machine knitting of garments and textile products, and producing crocheted garments and textile products to commercial order. The buyer sets the task and the accept criteria; Firsthand handles contributors, capture, QA, consent and delivery.

The data is human capture by paid, consented contributors, not robot or teleoperation data. This page is about knitting robot training data in the sense of what a policy can learn from watching skilled yarn-craft work, and it makes no claim that any robot has been trained on it or that any robot knits or crochets. People reach this topic through several phrasings, such as stereo crochet video, an egocentric yarn-craft dataset or custom textile arts data collection at scale. They describe one request from different angles.

Two. In the Textile Arts & Stitching Studio environment of the Creative Workshops & Crafts sector the catalog lists "Knitting": "Hand and machine knitting of garments and textile products." and "Crochet": "Producing crocheted garments and textile products to commercial order." Both are rated Medium, both list the capture setup Phone (head-mounted) / multiview rig, and both have the status Available on request. The first table below reproduces them exactly, with two nearby rows as context only.

Available on request means the work is scoped and recruited to spec when a buyer asks for it. This page therefore claims no completed collection, no delivered hours and no delivered stereo footage for either row. The studio holds nine rows in all, six Ready to deploy and three Available on request, and the three on request are Crochet, Embroidery and Knitting. The environment examples are hand embroidery, tailoring studio, quilting, needlepoint, custom garment fitting and hemming, so knitting and crochet are rows in the studio rather than among its listed examples.

Not stated. The catalog lists a head-mounted phone or a multiview rig for both rows and no stereo pair. Stereo applies here only through the general stereo offer: the site describes a stereo head rig and a hard-environments stack, and a stereo pair for a knitting or crochet order is a choice made in the skill spec and confirmed on a Spec Pilot.

This is why the page is written as a stereo and container page for two rows rather than as a description of existing stereo yarn-craft footage. If you need stereo, say so in the brief; if you are happy with the catalog setup, the egocentric knitting and crochet page describes that non-stereo route.

This page is stereo capture of hand and machine knitting and crochet. The egocentric knitting and crochet page covers the same two rows as first-person video without stereo as its subject. The egocentric quilting and needlepoint page covers two other studio rows, Quilting and Needlepoint, and is not stereo either.

The stereo embroidery and tailoring page is the nearest sibling and the one to read alongside this page. It treats all nine studio rows in stereo, Knitting and Crochet among them, with hand stitching, hemming, tailoring and quilting as its main subject. This page narrows to the two yarn-craft rows. The stereo garment sewing and fabric cutting page is a different environment, industrial sewing and cutting in Textile & Apparel Production. The second table below puts them side by side with their status as the site lists it.

The rows are written as whole jobs, not as step lists. Knitting is hand and machine work on garments and textile products, and Crochet is producing crocheted garments and textile products to commercial order. The site names no stitch, no casting on, no winding and no finishing step for either row, and it lists no standalone yarn winding or yarn handling row, so this page claims none.

What a spec can do is cut the job into steps the buyer names and give each a start and an end. Action segments carry start_ns, end_ns, verb and noun, so one piece of work becomes a sequence of labeled spans instead of one long clip, and contact events mark when a hand meets the work. The knitting row covers both hand and machine work in one description. A spec that wants only the hand part should say so, and one that wants a machine pass should name it. The site states no machine, no machine speed and no hand technique for the rows, and how a moving machine part is handled by a stereo pair is Not stated.

The site describes two stereo configurations. The stereo head rig, dataset 06, is a dual fisheye pair at 1080p per eye on a 62 mm baseline, with per-unit calibration, a 9-axis IMU and hand-tracking output, shipped as MCAP. The hard-environments stack is a 4K/60 head camera with a 120° lens, two 1080p/60 wrist cameras, depth at 848×480 and 30 fps and a 200 Hz 6-DoF IMU. Which of these a knitting or crochet order uses is Not stated and is agreed in the skill spec.

Whichever rig is chosen, calibration is measured every session and shipped with the episode, with a 0.28 px RMS reprojection gate. Hand pose is 21 joints per hand with a per-joint visibility flag, and the language data is one task instruction per episode. The site lists two capture modes: a head-mounted phone records the first-person view of real work, and a multiview rig with handheld grippers records a staged working set. Both rows list one setup that covers both, so whether a batch is real work or a staged set is settled in the spec. The studio, home or shop a collection runs in is Not stated.

Hand occlusion is the first case. The tool, the strand and the fingers hide one another, and a closing hand hides joints. The site states that the hands occlude the work and that fingers disappear behind objects. The per-joint visibility flag on the 21-joint hand pose marks joints that are not visible instead of guessing them, contact events mark when a hand meets the work, and a spec can ask for wrist or multiview angles so a grip the head camera loses is still recorded. The site names no needle, hook or yarn for these rows, so which tool the spec follows is set per program.

Thin, low-contrast strands and fine two-handed motion are the second case. The site says metric depth on sub-centimetre parts is flagged rather than interpolated, and that two-handed fine motion and small objects are where occlusion bites. Whether a strand of yarn counts as such a part, and how often it leaves the frame, is Not stated, so a Spec Pilot checks it on real footage.

Deformable fabric is the third. The site states that masks for deformable objects are shipped per frame rather than propagated, with instance masks stored at the depth resolution of 848×480. What counts as the object, the piece, the strand or the tool, is a spec decision. Lighting is the fourth: the site states no standard lighting values for these rows, so any lighting conditions are fixed in the skill spec for the place in question.

Every batch is validated against signed accept criteria before it ships, with a reject log. The site states that an episode is rejected when the manipulated object is out of frame for more than 15 percent of it, when more than 35 percent of depth pixels in the manipulation region are invalid, when a sync breach runs beyond the ceiling for more than three consecutive frames, when an undocumented bystander appears, or when the episode does not match the named skill spec. Episodes that fail are rejected and logged per batch, not delivered with a caveat.

Billing is per validated hour, and rejected items are recaptured at our cost. The site does not state an accuracy figure for step labels, and it lists no label for stitch count, tension, gauge, a dropped stitch or whether a piece is finished, so what counts as a pass for a step boundary is set in the accept criteria.

Firsthand offers MCAP delivery. MCAP is an open-source container format with its own official site at mcap.dev, a specification at mcap.dev/spec and a repository at github.com/foxglove/mcap; read those for what the container is and how it stores records, timestamps and attachments. This page describes the container only from those sources and does not state anything more about it.

The streams the site already states for a stereo capture are the stereo pair, depth with confidence maps, 21-joint hand pose, IMU, contact events, action segments, per-frame instance masks and calibration, with one task instruction per episode. Topic names, message schemas, encoding and rates are Not stated on the site and are agreed in the skill spec. Only dataset 06 is stated as shipped as MCAP; for any custom order, MCAP can be the only delivery or sit beside RLDS, LeRobot, WebDataset, HDF5 or zarr.

Yarn type or brand, needle or hook size, machine model, gauge, stitch counts, patterns, garment types, garment counts, task times, error rates, force or tactile data, a task-completion label and any safety or compliance certification are not stated on the site as standard fields. No studio, shop, client or brand is named for either row.

If your policy needs any of these, they are written into the skill spec as annotation depth and covered in the quote before capture begins. This page gives no needle, machine or other equipment-safety instructions.

The site’s off-the-shelf catalog lists about 324,800 hours across 18 datasets, including about 64,800 hours of stereo video in 9 datasets. None of them names crochet. One stereo set, S4 (2,984 hours; yarn mill, textile and garment; head-mounted stereo with synced left and right RGB, IMU and audio), lists stitching and knitting among its tasks. Whether S4 holds hand or machine knitting of garments is Not stated, so it is a way to look at the format and rig, not coverage of the Knitting row.

Stereo knitting and crochet collection to a spec is therefore custom collection: 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. Both rows are Available on request, so the scoped-to-spec timeline applies rather than the ready-now case, and the quote states it up front.

Every episode links to a signed contributor release through a documented consent chain, and the consent record is part of the delivery. Faces, plates and screens are detected and irreversibly blurred at ingest, and a reviewer confirms the result before anything ships, with the anonymization record delivered alongside. Footage that cannot be de-identified is discarded, and an undocumented bystander is a stated reject criterion. Anything beyond that, such as signage or labels on yarn, is not listed as a blur category, so raise it in the spec if it matters.

The buyer owns the dataset under a perpetual, irrevocable, non-exclusive-by-default licence, with exclusivity available. Consent is recorded per session. The executed agreement controls.

Describe the row you want, the steps you expect to see and where each starts and ends. Say whether one episode is a single step, a single piece or a run of work, whether the knitting is hand, machine or both, what counts as the object, what counts as a reject, and whether you want a stereo pair, wrist views or a multiview rig. 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.

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 last table below shows what each spec element decides.