L1 ⚛ L1-078 ⊙ Testnet

Event Camera (DVS) — intensity reconstruction from asynchronous events

Computational Photography · Asynchronous neuromorphic imaging

⬇ Download bundle 🚀 Submit to mainnet — earn 5% royalty

🎁 Contribute & earn PWM

Improve L1-078 — submit with or without uploading. Accepted submissions earn the 1/16 modify share of A(t).

✨ Submit without uploading free check · 1×

One click — no files, no typing. We submit a check of L1-078 for you; it goes through review and earns the 1/16 share once. One free no-upload Principle per account.

You'll be asked to log in first — then it's one click, no files.

⬆ Or submit by uploading your files

Upload your .md + .json as a modification of L1-078. Unlimited.

🛠 Helper: build your files in ChatGPT / Gemini / Claude

Layer template

⬇ principle.md ⬇ principle.json

Fork L1-078

⬇ L1-078.md ⬇ L1-078.json

Prompt — copy into your LLM

You are helping me submit a MODIFICATION of L1-078 (Event Camera (DVS) — intensity reconstruction from asynchronous events) to the PWM Protocol — a Principle (L1) artifact.

I will paste a Markdown template (or the current L1-078.md).
1. Rewrite the Markdown so the science is correct and clearly explained for my change.
2. Regenerate the sibling L1-078.json so EVERY field matches the Markdown.
3. Keep the schema in the "File Mapping" footer at the bottom of the MD.
4. Keep the parent reference unchanged unless I ask otherwise.
Rules: the Markdown is the source of truth; use SI units; do NOT invent benchmark scores.
Output each file in its own fenced code block tagged with the filename.

Here is my template:
[PASTE THE .md HERE]

⚙ Forward Model

y = `S.stream.asynchronous` `D.threshold.delta` `L.log_intensity` x

world state x (2D plus time video)

→

L · log intensity `L.log_intensity`

→

D · threshold · delta `D.threshold.delta`

→

S · stream · asynchronous `S.stream.asynchronous`

→

observation y (photon detector)

Noise: asynchronous event

Markdown — human-readable source of truth

⚙ auto-generated

⬇ L1-078.md

# ⚛ L1 Principle — Event Camera (DVS) — intensity reconstruction from asynchronous events **ID:** `L1-078` · **Status:** ⊙ Testnet (genesis catalog) > **🌐 Domain:** Computational Photography — *Asynchronous neuromorphic imaging* > **🎯 Problem class:** nonlinear inverse · **🧮 Solution space:** 2D plus time video > **📡 Carrier:** photon · **🌫 Noise:** asynchronous event > **⚖ Difficulty (δ):** 5 · **⛓ Block:** 41554171 --- ## 🧠 1. Introduction **Event Camera (DVS) — intensity reconstruction from asynchronous events** is a **nonlinear inverse problem** whose unknown lives in **2D plus time video** space, within the **Asynchronous neuromorphic imaging** sub-domain of **Computational Photography**. Measurements consist of photons collected by an optical detector via a **asynchronous log intensity change** sensing mechanism. The forward operator applies, in order: L · log intensity operator; D · threshold · delta operator; S · stream · asynchronous operator. Observations are corrupted by asynchronous event. Intensity reconstruction is ill-posed (DC level is unobservable; only log-differentials are sampled). Unique reconstruction requires an anchor frame (DAVIS hybrid output) or a strong prior (video smoothness, learned models). Contrast-threshold variance sigma_C and noise events add strong instability at low event rates. ## ⚙ 2. Forward Model Physical chain: **x** → L · log intensity → D · threshold · delta → S · stream · asynchronous → **y** (detector). ``` y = `S.stream.asynchronous` `D.threshold.delta` `L.log_intensity` x ``` **Measurement DAG:** | Primitive | What it does | |---|---| | `L.log_intensity` | L · log intensity operator | | `D.threshold.delta` | D · threshold · delta operator | | `S.stream.asynchronous` | S · stream · asynchronous operator | ## 🔬 3. Physics Fingerprint | Property | Value | |---|---| | Domain | Computational Photography | | Sub domain | Asynchronous neuromorphic imaging | | Carrier | photon | | Problem class | nonlinear_inverse | | Solution space | 2D_plus_time_video | | Noise model | asynchronous_event | | Integration axis | temporal | | Difficulty delta | 5 | | L dag | 2.8 | ## 📡 4. Measurement Model Intensity reconstruction is ill-posed (DC level is unobservable; only log-differentials are sampled). Unique reconstruction requires an anchor frame (DAVIS hybrid output) or a strong prior (video smoothness, learned models). Contrast-threshold variance sigma_C and noise events add strong instability at low event rates. | Metric | Value | |---|---| | Metric | PSNR_dB | | Secondary | SSIM | ## 📏 5. Operating Range (Ω) **Center problem class:** `event_to_intensity_reconstruction` · **Forward operator:** `dvs_log_intensity_threshold` **Center point:** | Parameter | Unit | Value | |---|---|---| | H | px | 640 | | W | px | 480 | | N events per sec | — | 1000000 | | Noise event rate | — | 0 | | Contrast threshold c | — | 0.2 | | Reconstruction rate hz | Hz | 60 | | Contrast threshold variance | — | 0 | **Allowed bounds:** | Parameter | Unit | Range | |---|---|---| | H | px | 240 – 1280 | | W | px | 180 – 960 | | Hot pixels | — | 0 – 100 | | N events per sec | — | 10000 – 100000000 | | Noise event rate | — | 0.0 – 100000 | | Contrast threshold c | — | 0.05 – 0.5 | | Reconstruction rate hz | Hz | 15 – 240 | | Refractory period drift | — | 0.0 – 1000 | | Contrast threshold variance | — | 0.0 – 0.1 | ## 🎯 6. Tolerance (ε) **Center tolerance:** 22.0 dB PSNR | Metric | Range | |---|---| | Psnr db | 12.0 – 32.0 | ## ⚖ 7. Hardness Function Hardness scales as **`epsilon_fn`** on **PSNR_dB**, with κ = `4000` and δ = `5`. ## 💾 8. Reference Dataset - **primary** · weight 1.0 · IPFS _(not pinned yet)_ ## 9. On-chain Registration - **Chain hash:** `0xa72e0d5ef4f3d0bd8dc596c336dff5a8be897f9205125c6c047f183679216e39` - **Chain tx hash:** `0xbbdefcb2d434660bc269d96a614c31e6c00cb029ca09aaee6f9722bcd4019d72` - **Chain block:** `41554171` --- ## File Mapping This bundle consists of: `L1-078.md`, `L1-078.json`. | File | Role | How to regenerate | |------|------|-------------------| | `L1-078.md` | Source of truth — edit this | Human or LLM | | `L1-078.json` | Structured metadata for the registry | LLM regenerates from the sections above | **Prompt for your LLM after editing this Markdown:** > Read the attached Markdown. Regenerate the sibling `.json` so every field matches. > Preserve the schema documented in the rows above. > Output each file in its own fenced code block tagged with the filename. > Output only the JSON object. _This Markdown was auto-synthesized from the catalog row for `L1-078`._ _Edit it, regenerate the JSON, and submit at [/submit](/submit) to claim the artifact._

💬

Ask the PWM AI agent AI

Ask about this Event Camera (DVS) — intensity reconstruction from asynchronous events — or how to improve it & earn PWM

Free to search; log in to chat with the agent & earn PWM. How it works ↗

⌥ Hierarchy

Upstream lineage and downstream children

↓ L2 Digital Twin 1 child ▾

L2-078-001 Event Camera Intensity Reconstruction Mismatch-Only Spec

⬇ Download bundle

2 files · click to download

📋 JSON metadata

{
  "artifact_id": "L1-078",
  "chain_block": 41554171,
  "chain_hash": "0xa72e0d5ef4f3d0bd8dc596c336dff5a8be897f9205125c6c047f183679216e39",
  "chain_tx_hash": "0xbbdefcb2d434660bc269d96a614c31e6c00cb029ca09aaee6f9722bcd4019d72",
  "domain": "Computational Photography",
  "hardness_fn": {
    "delta": 5,
    "kappa": 4000,
    "metric": "PSNR_dB",
    "type": "epsilon_fn"
  },
  "initiator_dataset": [
    {
      "ipfs_cid": null,
      "license_hash": null,
      "name": "primary",
      "weight": 1.0
    }
  ],
  "layer": "L1",
  "observable_profile": {
    "metric": "PSNR_dB",
    "regime": "Intensity reconstruction is ill-posed (DC level is unobservable; only log-differentials are sampled). Unique reconstruction requires an anchor frame (DAVIS hybrid output) or a strong prior (video smoothness, learned models). Contrast-threshold variance sigma_C and noise events add strong instability at low event rates.",
    "secondary": "SSIM"
  },
  "physics_fingerprint": {
    "L_DAG": 2.8,
    "carrier": "photon",
    "difficulty_delta": 5,
    "domain": "Computational Photography",
    "integration_axis": "temporal",
    "noise_model": "asynchronous_event",
    "primitives": [
      "L.log_intensity",
      "D.threshold.delta",
      "S.stream.asynchronous"
    ],
    "problem_class": "nonlinear_inverse",
    "sensing_mechanism": "asynchronous_log_intensity_change",
    "solution_space": "2D_plus_time_video",
    "sub_domain": "Asynchronous neuromorphic imaging",
    "title": "Event Camera (DVS) \u2014 intensity reconstruction from asynchronous events"
  },
  "size_tiers": {
    "allowed_forward_operators": [
      "dvs_log_intensity_threshold",
      "atis_time_contrast",
      "celex_fovea"
    ],
    "allowed_omega_dimensions": [
      "H",
      "W",
      "N_events_per_sec",
      "contrast_threshold_C",
      "reconstruction_rate_Hz",
      "contrast_threshold_variance",
      "noise_event_rate",
      "refractory_period_drift",
      "hot_pixels"
    ],
    "allowed_problem_classes": [
      "event_to_intensity_reconstruction",
      "event_based_deblur",
      "event_optical_flow"
    ],
    "center_spec": {
      "epsilon_fn_center": "22.0 dB PSNR",
      "forward_operator": "dvs_log_intensity_threshold",
      "input_format": "event_stream_plus_optional_anchor_frame",
      "omega": {
        "H": 640,
        "N_events_per_sec": 1000000,
        "W": 480,
        "contrast_threshold_C": 0.2,
        "contrast_threshold_variance": 0.0,
        "noise_event_rate": 0.0,
        "reconstruction_rate_Hz": 60
      },
      "problem_class": "event_to_intensity_reconstruction"
    },
    "epsilon_bounds": {
      "psnr_db": [
        12.0,
        32.0
      ]
    },
    "omega_bounds": {
      "H": [
        240,
        1280
      ],
      "N_events_per_sec": [
        10000,
        100000000
      ],
      "W": [
        180,
        960
      ],
      "contrast_threshold_C": [
        0.05,
        0.5
      ],
      "contrast_threshold_variance": [
        0.0,
        0.1
      ],
      "hot_pixels": [
        0,
        100
      ],
      "noise_event_rate": [
        0.0,
        100000
      ],
      "reconstruction_rate_Hz": [
        15,
        240
      ],
      "refractory_period_drift": [
        0.0,
        1000
      ]
    }
  },
  "staked_pwm": 0.0,
  "status": "testnet",
  "sub_domain": "Asynchronous neuromorphic imaging",
  "title": "Event Camera (DVS) \u2014 intensity reconstruction from asynchronous events"
}

🎁

Earn PWM by promoting this to mainnet

• PWM emission on accept — your share of A(t) (1/16 modify · 3/16 new)
• 5% upstream royalty on every downstream L4 mining event, forever
• Your wallet is enrolled the moment the founder accepts

🚀 Go to submit ↑

How: use the “Contribute & earn PWM” panel at the top — either Submit without uploading (one click) or upload your files. Then the review runs and, on accept, it’s promoted to mainnet.