feat: frame/time pipeline split + output validation

- Add PipelineType enum + pipeline() to ProcessorType - Split ProcessorPool into frame_slots (max 2) and time_slots (max 1) - Add can_start_for() for pipeline-aware scheduling - Add validate_output_file() — checks JSON validity before marking complete - Add 3 unit tests for validate_output_file() - Create DESIGN/FRAME_TIME_PIPELINE_V1.0.md (492 lines)
2026-05-23 21:14:28 +08:00
parent dddb5d4cbd
commit f8bcc0356c
5 changed files with 729 additions and 70 deletions
--- a/docs_v1.0/DESIGN/FRAME_TIME_PIPELINE_V1.0.md
+++ b/docs_v1.0/DESIGN/FRAME_TIME_PIPELINE_V1.0.md
@@ -0,0 +1,492 @@
+---
+title: "Frame / Time 雙產線分流協作設計 v1.0"
+version: "1.0"
+date: "2026-05-23"
+author: "M5"
+status: "draft"
+scope: "architecture, worker, storage"
+---
+
+# Frame / Time 雙產線分流協作設計 v1.0
+
+|  Scope  | Status | Applies to |
+|---------|--------|------------|
+| architecture / worker / storage | draft | momentry_core worker pipeline |
+
+---
+
+## 1. 緣起與問題
+
+### 1.1 現狀問題
+
+worker 將所有 processor 混在一起平行執行，導致：
+
+| 問題 | 說明 |
+|------|------|
+| OOM | `max_concurrent=6` 時 6 個 Python 行程同時載入模型 → 記憶體不足被 kill |
+| 資源競爭 | 多個 processor 各自開 ffmpeg decode 同一部影片 → 6 倍 decode |
+| 重試粒度粗 | 一個 processor 失敗 → 整部片全部重來 |
+| 進度不精確 | 0% → 100%，中間無細粒度進度 |
+
+### 1.2 手動 vs Worker 差異
+
+| 面向 | 手動執行 | Worker 自動化 |
+|------|---------|--------------|
+| 執行方式 | 循序，一次一個 processor | 平行，最多 max_concurrent 個 |
+| 產出檢查 | 人工確認 JSON 內容正確 | `output_path.exists()` 僅檢查檔案存在 |
+| 資源 | 單一模型在記憶體 | 多個模型競爭記憶體 |
+
+### 1.3 核心結論
+
+兩個根本問題必須解決：
+
+1. **產線分流** — Frame-base 與 Time-base processor 不應混合排程
+2. **Frame-level resource management** — 透過 MarkbaseFMS 統一 frame 存取
+
+---
+
+## 2. 雙產線架構
+
+### 2.1 Pipeline Overview
+
+```mermaid
+graph TD
+    Input[Input Video] --> Probe
+    Probe -->|frame info| FMS[MarkbaseFMS]
+    Probe -->|audio track| TimePipe[Time Pipeline]
+
+    FMS -->|frame batches| FramePipe[Frame Pipeline]
+    FMS -->|cache / align / convert| FMS
+
+    subgraph FramePipe [Frame Pipeline]
+        CUT[CUT - scene detection]
+        YOLO[YOLO - object detection]
+        Face[Face - face detection]
+        OCR[OCR - text detection]
+        Pose[Pose - pose estimation]
+    end
+
+    subgraph TimePipe [Time Pipeline]
+        ASR[ASR - speech recognition]
+        ASRX[ASRX - speaker diarization]
+    end
+
+    CUT --> YOLO
+    CUT --> Face
+    CUT --> OCR
+    CUT --> Pose
+
+    ASR --> ASRX
+
+    YOLO & Face & OCR & Pose --> Merge[Merge Processor Results]
+    ASR & ASRX --> Merge
+
+    Face --> Lip[Lip Sync]
+    ASR --> Lip
+    ASRX --> Lip
+
+    Merge -->|all essential done| PostProcess
+    Lip --> PostProcess
+
+    subgraph PostProcess [Post-processing]
+        R1[Rule 1 Chunking]
+        R3[Rule 3 Chunking]
+        TK[TKG Build]
+        W1[5W1H Agent]
+        ID[Identity Agent]
+    end
+```
+
+### 2.2 各產線定義
+
+#### Frame Pipeline（frame-based）
+
+| Processor | 輸入 | 輸出 | 瓶頸資源 | 產線 |
+|-----------|------|------|---------|------|
+| CUT | frame (降解析) | scene.json | CPU | Frame |
+| YOLO | frame batch | yolo.json | GPU | Frame |
+| Face | frame batch | face.json | ANE/GPU | Frame |
+| OCR | frame batch | ocr.json | CPU | Frame |
+| Pose | frame batch | pose.json | GPU | Frame |
+
+#### Time Pipeline（time-based）
+
+| Processor | 輸入 | 輸出 | 瓶頸資源 | 產線 |
+|-----------|------|------|---------|------|
+| ASR | audio stream | asr.json | GPU/CPU | Time |
+| ASRX | audio stream + ASR result | asrx.json | CPU | Time |
+
+#### 合流點
+
+| 項目 | 需要 | 產出 |
+|------|------|------|
+| Lip Sync | Face + ASR + ASRX | lip.json (who speaks when) |
+| Rule 1 Chunking | ASR + ASRX | sentence chunks |
+| Rule 3 Chunking | CUT + ASR | scene chunks |
+| TKG Build | 所有 processor | tkg_nodes / tkg_edges |
+| 5W1H Agent | CUT + ASR | story summary |
+| Identity Agent | Face + ASRX | identity bindings |
+
+---
+
+## 3. MarkbaseFMS 底層設計
+
+### 3.1 三層架構
+
+```
+                    Application
+    YOLO / Face / OCR / Pose / CUT
+    讀取 frame buffer → 直接做 inference
+                    │
+                    │ frame-aligned access
+                    ▼
+┌───────────────────────────────────┐
+│  FMS Filesystem Layer             │
+│  Layout: frame data 連續存放      │
+│  不跨 frame split                 │
+│  格式: opaque raw buffer          │
+│  metadata: 獨立區域               │
+│  read-ahead: 預取下一個 Block     │
+│  alignment: 每 frame page-aligned │
+└──────────────────┬────────────────┘
+                    │ page-aligned (4096)
+┌──────────────────▼────────────────┐
+│  FMS Cache Layer                  │
+│  unit: FrameBlock (64 frames)     │
+│  alignment: page boundary/frame   │
+│  eviction: LRU on Block           │
+│  pin: 使用中的 frame 不 evict     │
+│  prefetch: 預拉下一個 Block       │
+│  Direct I/O bypass OS cache       │
+└──────────────────┬────────────────┘
+                    │ block-aligned (4K / 64K)
+┌──────────────────▼────────────────┐
+│  Block Device / Storage           │
+│  sector alignment 4K              │
+│  FrameBlock = N 個連續 sectors    │
+│  無跨 sector split                │
+│  atomic write per block           │
+│  O_DIRECT 直接 IO                 │
+└───────────────────────────────────┘
+```
+
+### 3.2 對齊原則
+
+#### 各層對齊要求
+
+| 層級 | 對齊單位 | 原因 |
+|------|---------|------|
+| Block Device | 4K sector | 現代 SSD 原生 sector，避免 RMW |
+| Cache | 4096 page | `mmap` + `madvise` 大頁面，減少 TLB miss |
+| Filesystem | frame block (64 frames) | 連續 layout，預測性 read-ahead |
+| Frame Buffer | 16 bytes stride | NEON SIMD，MPS/ANE texture 要求 |
+| Block Index | power of 2 | index 用 bit shift + mask，無需除法 |
+
+#### Frame Buffer Layout
+
+```
+每個 frame 的 raw buffer：
+
+  ┌──────────────┐
+  │  Y Plane     │  ← 16-byte aligned stride, page-aligned offset
+  │  (width×h)   │
+  ├──────────────┤
+  │  UV Plane    │  ← 16-byte aligned stride
+  │  (w/2×h/2×2) │      (NV12 interleaved)
+  └──────────────┘
+
+   frame buffer offset:   page-aligned (4096)
+   row stride:            align(width * pixel_size, 16)
+   frame size:            align(total_bytes, page_size)
+```
+
+### 3.3 Block 排列方式
+
+```
+每個 Block 包含連續 64 frames（configurable）：
+
+  Block index ≤— bit shift (frame_num / 64)
+  Frame offset = base + (frame_num & 63) * frame_size  ← bit mask
+
+Block Dispatching Strategy:
+
+  Worker 要求 "batch 0, format=RGB, width=640"
+  FMS:
+    1. Check Block Cache (RAM):
+         Block 0 frames 0-63 是否已 decode?
+         ✅ hit → 直接回傳
+         ❌ miss → decode 64 frames → 存入 cache → 回傳
+    2. Format Conversion (on-the-fly):
+         原始 NV12 → per request:
+           YOLO → RGB (float32 normalized)
+           Face → RGB (uint8)
+           OCR  → Gray (uint8)
+           Pose → RGB (uint8)
+           CUT  → RGB (降解析, uint8)
+```
+
+### 3.4 儲存 Layout
+
+```
+Disk Layout (per file_uuid):
+
+┌──────────────────────────────────────────────┐
+│              Metadata Region                  │
+│  - file_uuid       (32 bytes)                │
+│  - total_frames    (u32)                     │
+│  - width, height   (u32 × 2)                 │
+│  - fps             (f64)                     │
+│  - pixel_format    (u8 : 0=NV12)             │
+│  - block_capacity  (u32, default 64)         │
+│  - block_count     (u32)                     │
+│  - frame_size      (u32, bytes per raw frame)│
+│  - block_offsets   [u64 × block_count]       │
+│  Padding to 4096                             │
+├──────────────────────────────────────────────┤
+│              Data Region                      │
+│  Block 0:  frames [0, 63]                    │
+│    frame_0:  [frame_size bytes]  ← page align│
+│    frame_1:  [frame_size bytes]               │
+│    ...                                       │
+│    frame_63: [frame_size bytes]              │
+│  Block 1:  frames [64, 127]                  │
+│  ...                                         │
+│  Block N:  frames [N*64, ...]               │
+└──────────────────────────────────────────────┘
+```
+
+### 3.5 記憶體管理
+
+```
+┌────────────────────────────────────────────┐
+│              FMS Cache                      │
+├──────────┬──────────┬──────────┬───────────┤
+│ Block 0  │ Block 1  │ Block 2  │  ...      │ ← mmap'd
+│ (64 fr)  │ (64 fr)  │ (64 fr)  │           │   or anonymous
+├──────────┴──────────┴──────────┴───────────┤
+│ LRU eviction policy                         │
+│ max_memory = configurable                   │
+│   (default: 256 frames ~1.5GB @1080p NV12)  │
+│ pin_count: 正在被 processor 存取的 frame    │
+│   pinned frame 不參與 LRU eviction           │
+└─────────────────────────────────────────────┘
+```
+
+---
+
+## 4. Worker 調度器修改
+
+### 4.1 Processor 產線標記
+
+```rust
+enum PipelineType {
+    Frame,   // frame-based
+    Time,    // time-based  
+    Cross,   // needs both frame + time
+}
+
+impl ProcessorType {
+    fn pipeline(&self) -> PipelineType {
+        match self {
+            Self::Cut
+            | Self::Yolo
+            | Self::Face
+            | Self::Ocr
+            | Self::Pose => PipelineType::Frame,
+
+            Self::Asr | Self::Asrx => PipelineType::Time,
+
+            Self::Story
+            | Self::Tkg
+            | Self::Identity
+            | Self::FiveW1h
+            | Self::Caption => PipelineType::Cross,
+        }
+    }
+}
+```
+
+### 4.2 資源配額
+
+| 產線 | Max Concurrent | 策略 |
+|------|---------------|------|
+| Frame | 2 | 最多同時 2 個 frame processor |
+| Time | 1 | 一次只跑 1 個 audio processor |
+| Cross | 1 | Frame + Time 都完成後才允許 |
+
+Frame pipeline 內部建議順序：
+
+```
+CUT (先確定場景邊界)
+  → YOLO / Face 可同時 (GPU-bound)
+  → OCR / Pose 可同時 (CPU/GPU mixed)
+```
+
+### 4.3 產出驗證加強
+
+```rust
+// 目前 (job_worker.rs:346):
+if output_path.exists() {
+    mark_completed();
+}
+
+// 改為:
+if output_path.exists() {
+    match validate_processor_output(&output_path, processor_type) {
+        Ok(true) => mark_completed(),
+        Ok(false) => retry_or_fail(),
+        Err(e) => mark_failed(e),
+    }
+}
+
+fn validate_processor_output(path: &Path, pt: ProcessorType) -> Result<bool> {
+    let content = std::fs::read_to_string(path)?;
+    let json: serde_json::Value = serde_json::from_str(&content)?;
+    // 至少要有基本欄位
+    match pt {
+        ProcessorType::Asr => json.get("segments").is_some(),
+        ProcessorType::Yolo => json.get("frames").is_some(),
+        ProcessorType::Face => json.get("frames").is_some(),
+        // ...
+    };
+    Ok(true) // or false
+}
+```
+
+### 4.4 啟動順序
+
+```
+1. Probe → 決定 frame 數、audio 格式
+2. CUT (Frame Pipeline 第一階段—決定場景邊界)
+3. Frame Pipeline 平行: YOLO / Face / OCR / Pose
+   Time Pipeline 平行: ASR
+4. ASRX (依賴 ASR 結果)
+5. Lip Sync (等待 Face + ASR + ASRX)
+6. 所有 processor 完成 → 合流：
+   - Rule 1 / Rule 3 Chunking
+   - Face Trace / TKG
+   - 5W1H / Identity Agent
+```
+
+---
+
+## 5. Processor 串接 FMS
+
+### 5.1 FMS API
+
+```rust
+// Frame access API (async)
+impl FmsClient {
+    /// 取得單一 frame buffer
+    async fn get_frame(
+        &self,
+        file_uuid: &str,
+        frame_num: u32,
+        format: PixelFormat,
+        width: u32,
+        height: u32,
+    ) -> Result<RawFrame>;
+
+    /// 取得一個 batch frames (block-aligned)
+    async fn get_block(
+        &self,
+        file_uuid: &str,
+        block_idx: u32,
+        format: PixelFormat,
+        width: u32,
+        height: u32,
+    ) -> Result<FrameBlock>;
+
+    /// 串流 frames (lazy batch iteration)
+    fn stream_frames(
+        &self,
+        file_uuid: &str,
+        range: Range<u32>,
+        format: PixelFormat,
+        width: u32,
+        height: u32,
+    ) -> FrameStream;
+}
+```
+
+### 5.2 YOLO 為例：processor 修改
+
+```
+目前:
+  processor::process_yolo(video_path, output_path, uuid)
+    → 自己開 ffmpeg decode
+    → 逐 frame 處理
+    → 寫入 yolo.json
+
+改為 FMS-based:
+  processor::process_yolo_with_fms(fms, uuid, output_path)
+    → let results = Vec::new()
+    → for batch in fms.stream_frames(uuid, 0..total, RGB, 640, 640):
+    →     let detections = yolo_model.infer(batch)
+    →     results.push(detections)
+    →     write partial → yolo.partial.{batch_idx}.json
+    → merge_partials() → yolo.json
+```
+
+### 5.3 Partial Result Merge
+
+```
+Frame Pipeline 產出多個 partial JSON：
+
+  yolo.0000.json    (frames 0-63)
+  yolo.0001.json    (frames 64-127)
+  ...
+  yolo.0063.json    (frames 4032-4095)
+
+Worker 合併為單一 yolo.json：
+
+```rust
+async fn merge_partials(
+    uuid: &str,
+    processor: &str,
+    partial_dir: &Path,
+    output_path: &Path,
+) -> Result<()> {
+    let partials = read_sorted_partials(uuid, processor, partial_dir);
+    let merged = merge_detections(partials);
+    write_json(output_path, merged)
+}
+```
+```
+
+---
+
+## 6. 實作優先序
+
+| 優先 | 項目 | 說明 |
+|------|------|------|
+| P0 | Worker 產線分流 | ProcessorType::pipeline() + frame_slots / time_slots 分開計算 |
+| P0 | 產出驗證加強 | `output_path.exists()` + JSON validity + schema check |
+| P1 | FMS FrameBlock 資料結構 | 含 16-byte stride / 4096 page alignment |
+| P1 | mmap-based frame cache | page-aligned frame buffers, LRU eviction |
+| P2 | FMS API 實作 | get_frame / get_block / stream_frames |
+| P2 | YOLO processor 串接 FMS | 改為 stream_frames 方式 |
+| P2 | 其他 processor 串接 FMS | Face / OCR / Pose / CUT |
+| P3 | FMS Direct I/O + sector alignment | O_DIRECT bypass OS page cache |
+| P3 | Prefetch / readahead | 預測下一個 block 並提前載入 |
+
+---
+
+## 7. 注意事項
+
+| # | 項目 |
+|---|------|
+| 1 | raw buffer 格式依 processor 需求轉換，FMS 負責 NV12 → RGB / Gray |
+| 2 | Time pipeline 的 ASR/ASRX **不經過 FMS**，直接處理 audio stream |
+| 3 | macOS 的 `mmap` 支援 page-aligned，但 `O_DIRECT` 需確認 compat |
+| 4 | FrameBlock size (64) 可配置，但需維持 power-of-2 |
+| 5 | FMS 只管理 frame lifecycle，不處理 processor-specific 邏輯 |
+| 6 | 多 processor 共享 frame 時，FMS 保證只 decode 一次 |
+
+---
+
+## 版本歷史
+
+| Version | Date | Author | Changes |
+|---------|------|--------|---------|
+| 1.0 | 2026-05-23 | M5 | 初版 |
--- a/src/core/db/mod.rs
+++ b/src/core/db/mod.rs
@@ -43,7 +43,7 @@ pub use mongodb_db::MongoDb;
 pub use postgres_db::{
    Bm25Result, CandidateRecord, CreateApiKeyConfig, FileIdentityRecord, FileRecord,
    HybridSearchResult, IdentityChunkRecord, IdentityDetailRecord, IdentityFaceRecord,
-    IdentityFileRecord, MonitorJob, MonitorJobStats, MonitorJobStatus, PostgresDb,
+    IdentityFileRecord, MonitorJob, MonitorJobStats, MonitorJobStatus, PipelineType, PostgresDb,
    ProcessorJobStatus, ProcessorResult, ProcessorType, ResourceRecord, VideoRecord, VideoStatus,
 };
 pub use qdrant_db::{QdrantDb, VectorPayload};
--- a/src/core/db/postgres_db.rs
+++ b/src/core/db/postgres_db.rs
@@ -559,6 +559,31 @@ impl ProcessorType {
            ProcessorType::FiveW1H,
        ]
    }
+
+    /// Pipeline type for scheduling: Frame-based, Time-based, or Cross (needs both).
+    pub fn pipeline(&self) -> PipelineType {
+        match self {
+            Self::Cut
+            | Self::Yolo
+            | Self::Face
+            | Self::Ocr
+            | Self::Pose
+            | Self::VisualChunk
+            | Self::Scene => PipelineType::Frame,
+
+            Self::Asr | Self::Asrx => PipelineType::Time,
+
+            Self::Story | Self::FiveW1H => PipelineType::Cross,
+        }
+    }
+}
+
+/// Pipeline classification for worker scheduling.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub enum PipelineType {
+    Frame,
+    Time,
+    Cross,
 }

 #[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq)]
--- a/src/worker/job_worker.rs
+++ b/src/worker/job_worker.rs
@@ -12,8 +12,8 @@ use crate::core::chunk::{rule1_ingest, rule3_ingest};
 use crate::core::config::OUTPUT_DIR;
 use crate::core::db::qdrant_db::QdrantDb;
 use crate::core::db::{
-    schema, MonitorJobStatus, PostgresDb, ProcessorJobStatus, RedisClient, VectorPayload,
-    VideoStatus,
+    schema, MonitorJobStatus, PipelineType, PostgresDb, ProcessorJobStatus, ProcessorType,
+    RedisClient, VectorPayload, VideoStatus,
 };
 use crate::core::embedding::Embedder;
 use crate::core::processor::heuristic_scene::generate_scene_meta;
@@ -338,62 +338,81 @@ impl JobWorker {
                .await?;

            // Check if output file already exists on disk (source of truth)
+            // and validate that it's a parseable JSON with expected structure.
            let output_path = PathBuf::from(OUTPUT_DIR.as_str()).join(format!(
                "{}.{}.json",
                job.uuid,
                processor_type.as_str()
            ));
            if output_path.exists() {
-                info!(
-                    "Processor {} output file exists, marking completed and skipping",
-                    processor_type.as_str()
-                );
-                self.db
-                    .update_processor_progress(
-                        &job.uuid,
-                        processor_type.as_str(),
-                        total_frames,
-                        total_frames,
-                        "completed",
-                    )
-                    .await?;
-                let total = total_frames as i32;
-                self.redis
-                    .update_worker_processor_status(
-                        &job.uuid,
-                        processor_type.as_str(),
-                        "completed",
-                        None,
-                        total,
-                        total,
-                        0,
-                        0,
-                        0,
-                    )
-                    .await?;
-                started_count += 1;
-                // 覆寫 result_map 讓相依性檢查能正確判斷
-                result_map.insert(
-                    *processor_type,
-                    crate::core::db::ProcessorResult {
-                        id: 0,
-                        job_id: job.id,
-                        processor_type: *processor_type,
-                        status: ProcessorJobStatus::Completed,
-                        started_at: None,
-                        completed_at: None,
-                        duration_secs: None,
-                        chunks_produced: 0,
-                        frames_processed: total_frames as i32,
-                        output_size_bytes: 0,
-                        error_message: None,
-                        output_data: None,
-                        retry_count: 0,
-                        created_at: String::new(),
-                        updated_at: String::new(),
-                    },
-                );
-                continue;
+                match validate_output_file(&output_path, *processor_type) {
+                    Ok(true) => {
+                        info!(
+                            "Processor {} output file exists and valid, marking completed",
+                            processor_type.as_str()
+                        );
+                        self.db
+                            .update_processor_progress(
+                                &job.uuid,
+                                processor_type.as_str(),
+                                total_frames,
+                                total_frames,
+                                "completed",
+                            )
+                            .await?;
+                        let total = total_frames as i32;
+                        self.redis
+                            .update_worker_processor_status(
+                                &job.uuid,
+                                processor_type.as_str(),
+                                "completed",
+                                None,
+                                total,
+                                total,
+                                0,
+                                0,
+                                0,
+                            )
+                            .await?;
+                        started_count += 1;
+                        result_map.insert(
+                            *processor_type,
+                            crate::core::db::ProcessorResult {
+                                id: 0,
+                                job_id: job.id,
+                                processor_type: *processor_type,
+                                status: ProcessorJobStatus::Completed,
+                                started_at: None,
+                                completed_at: None,
+                                duration_secs: None,
+                                chunks_produced: 0,
+                                frames_processed: total_frames as i32,
+                                output_size_bytes: 0,
+                                error_message: None,
+                                output_data: None,
+                                retry_count: 0,
+                                created_at: String::new(),
+                                updated_at: String::new(),
+                            },
+                        );
+                        continue;
+                    }
+                    Ok(false) => {
+                        warn!(
+                            "Processor {} output file exists but content invalid, will reprocess",
+                            processor_type.as_str()
+                        );
+                        // fall through → reprocess
+                    }
+                    Err(e) => {
+                        warn!(
+                            "Processor {} output validation error: {}, will reprocess",
+                            processor_type.as_str(),
+                            e
+                        );
+                        // fall through → reprocess
+                    }
+                }
            }

            // Check if processor already in terminal state
@@ -484,10 +503,15 @@ impl JobWorker {
                }
            }

-            // Check capacity before starting processor
-            if !self.processor_pool.can_start().await {
+            // Check pipeline capacity before starting processor
+            if !self.processor_pool.can_start_for(processor_type.pipeline()).await {
                info!(
-                    "Max concurrent processors reached, skipping remaining processors for job {}",
+                    "Max {} processors reached, skipping remaining processors for job {}",
+                    match processor_type.pipeline() {
+                        PipelineType::Frame => "frame",
+                        PipelineType::Time => "time",
+                        PipelineType::Cross => "cross",
+                    },
                    job.uuid
                );
                // 為所有未啟動的 processors 創建 Skipped 記錄
@@ -1180,6 +1204,35 @@ impl JobWorker {
    }
 }

+/// 驗證 processor 輸出檔案的完整性。
+/// 回傳 Ok(true) 表示有效，Ok(false) 表示檔案存在但內容異常需重跑，Err 表示檢查失敗。
+fn validate_output_file(path: &std::path::Path, processor_type: ProcessorType) -> Result<bool> {
+    let content = match std::fs::read_to_string(path) {
+        Ok(c) => c,
+        Err(_) => return Ok(false),
+    };
+    if content.trim().is_empty() {
+        return Ok(false);
+    }
+    let json: serde_json::Value = match serde_json::from_str(&content) {
+        Ok(v) => v,
+        Err(_) => return Ok(false),
+    };
+    // 依 processor type 檢查必要欄位
+    let valid = match processor_type {
+        ProcessorType::Asr => json.get("segments").and_then(|s| s.as_array()).map_or(false, |a| !a.is_empty()),
+        ProcessorType::Asrx => json.get("segments").and_then(|s| s.as_array()).map_or(false, |a| !a.is_empty()),
+        ProcessorType::Yolo => json.get("frames").and_then(|f| f.as_object()).is_some(),
+        ProcessorType::Face => json.get("frames").and_then(|f| f.as_object()).is_some(),
+        ProcessorType::Ocr => json.get("frames").and_then(|f| f.as_object()).is_some(),
+        ProcessorType::Pose => json.get("frames").and_then(|f| f.as_object()).is_some(),
+        ProcessorType::Cut => json.get("segments").or_else(|| json.get("scenes")).and_then(|s| s.as_array()).map_or(false, |a| !a.is_empty()),
+        // VisualChunk / Scene / Story / FiveW1H: 只檢查是 valid JSON 即可
+        _ => true,
+    };
+    Ok(valid)
+}
+
 #[cfg(test)]
 mod tests {
    use super::*;
@@ -1190,4 +1243,34 @@ mod tests {
        assert!(config.enabled);
        assert!(config.max_concurrent >= 1);
    }
+
+    fn test_validate_path(name: &str) -> std::path::PathBuf {
+        let dir = std::env::temp_dir().join(format!("test_validate_{}", name));
+        let _ = std::fs::create_dir_all(&dir);
+        dir.join("output.json")
+    }
+
+    #[test]
+    fn test_validate_output_empty() {
+        let path = test_validate_path("empty");
+        std::fs::write(&path, "").unwrap();
+        assert!(!validate_output_file(&path, ProcessorType::Yolo).unwrap_or(false));
+        let _ = std::fs::remove_file(&path);
+    }
+
+    #[test]
+    fn test_validate_output_invalid_json() {
+        let path = test_validate_path("invalid");
+        std::fs::write(&path, "not json").unwrap();
+        assert!(!validate_output_file(&path, ProcessorType::Yolo).unwrap_or(false));
+        let _ = std::fs::remove_file(&path);
+    }
+
+    #[test]
+    fn test_validate_output_yolo_ok() {
+        let path = test_validate_path("yolo_ok");
+        std::fs::write(&path, r#"{"frames":{"1":{"detections":[]}}}"#).unwrap();
+        assert!(validate_output_file(&path, ProcessorType::Yolo).unwrap_or(false));
+        let _ = std::fs::remove_file(&path);
+    }
 }
--- a/src/worker/processor.rs
+++ b/src/worker/processor.rs
@@ -7,35 +7,56 @@ use std::sync::Arc;
 use tokio::sync::{mpsc, RwLock};
 use tracing::{error, info, warn};

+
+
 /// Guard that ensures processor pool cleanup runs even if the task panics.
 struct ProcessorCleanupGuard {
    job_id: i32,
    running: Arc<RwLock<HashMap<i32, ProcessorHandle>>>,
    running_count: Arc<RwLock<usize>>,
+    frame_count: Arc<RwLock<usize>>,
+    time_count: Arc<RwLock<usize>>,
+    pipeline: PipelineType,
 }

 impl Drop for ProcessorCleanupGuard {
    fn drop(&mut self) {
        use tokio::sync::TryLockError;
-        // 嘗試同步清理；若 lock 被佔用則跳過（避免 deadlock）
        if let Ok(mut guard) = self.running.try_write() {
            guard.remove(&self.job_id);
        } else {
-            warn!("[ProcessorCleanupGuard] running lock contended, skipping cleanup");
+            warn!("[ProcessorCleanupGuard] running lock contended");
        }
        if let Ok(mut guard) = self.running_count.try_write() {
-            if *guard > 0 {
-                *guard -= 1;
+            if *guard > 0 { *guard -= 1; }
+        }
+        match self.pipeline {
+            PipelineType::Frame => {
+                if let Ok(mut guard) = self.frame_count.try_write() {
+                    if *guard > 0 { *guard -= 1; }
+                }
            }
-        } else {
-            warn!("[ProcessorCleanupGuard] running_count lock contended, skipping cleanup");
+            PipelineType::Time => {
+                if let Ok(mut guard) = self.time_count.try_write() {
+                    if *guard > 0 { *guard -= 1; }
+                }
+            }
+            PipelineType::Cross => {} // cross pipeline not tracked in slot counts
        }
    }
 }

+#[derive(Debug)]
+struct ProcessorHandle {
+    #[allow(dead_code)]
+    processor_type: ProcessorType,
+    cancel_tx: mpsc::Sender<()>,
+    child_pid: Arc<RwLock<Option<i32>>>,
+}
+
 use crate::core::config::{OUTPUT_DIR, PYTHON_PATH, SCRIPTS_DIR};
 use crate::core::db::{
-    MonitorJob, PostgresDb, ProcessorJobStatus, ProcessorType, QdrantDb, RedisClient,
+    MonitorJob, PipelineType, PostgresDb, ProcessorJobStatus, ProcessorType, QdrantDb, RedisClient,
 };
 use crate::core::processor;
 use crate::core::processor::asr::AsrResult;
@@ -67,19 +88,19 @@ pub struct ProcessorTask {
    pub frame_dir: Option<String>,
 }

+/// Frame pipeline max concurrent processors (hard limit).
+const FRAME_SLOT_MAX: usize = 2;
+/// Time pipeline max concurrent processors (audio is heavy, run 1 at a time).
+const TIME_SLOT_MAX: usize = 1;
+
 pub struct ProcessorPool {
    db: Arc<PostgresDb>,
    redis: Arc<RedisClient>,
    config_max: usize,
    running: Arc<RwLock<HashMap<i32, ProcessorHandle>>>,
    running_count: Arc<RwLock<usize>>,
-}
-
-struct ProcessorHandle {
-    #[allow(dead_code)]
-    processor_type: ProcessorType,
-    cancel_tx: mpsc::Sender<()>,
-    child_pid: Arc<RwLock<Option<i32>>>,
+    running_frame_count: Arc<RwLock<usize>>,
+    running_time_count: Arc<RwLock<usize>>,
 }

 impl ProcessorPool {
@@ -90,6 +111,8 @@ impl ProcessorPool {
            config_max: max_concurrent,
            running: Arc::new(RwLock::new(HashMap::new())),
            running_count: Arc::new(RwLock::new(0)),
+            running_frame_count: Arc::new(RwLock::new(0)),
+            running_time_count: Arc::new(RwLock::new(0)),
        }
    }

@@ -105,10 +128,27 @@ impl ProcessorPool {
        count < max
    }

+    /// 檢查特定產線是否可啟動新的 processor。
+    /// Frame pipeline 最多 FRAME_SLOT_MAX 個，Time pipeline 最多 TIME_SLOT_MAX 個。
+    pub async fn can_start_for(&self, pipeline: PipelineType) -> bool {
+        let count = *self.running_count.read().await;
+        let max = self.current_max().await;
+        if count >= max {
+            return false;
+        }
+        match pipeline {
+            PipelineType::Frame => *self.running_frame_count.read().await < FRAME_SLOT_MAX,
+            PipelineType::Time => *self.running_time_count.read().await < TIME_SLOT_MAX,
+            PipelineType::Cross => false, // cross pipeline = wait until frame+time done
+        }
+    }
+
    /// 清理 stale running state：若系統中實際運行的 processor 比記錄少，修正 count
    pub async fn sweep_stale(&self) {
        let handle_count = self.running.read().await.len();
        let count = *self.running_count.read().await;
+        let frame_count = *self.running_frame_count.read().await;
+        let time_count = *self.running_time_count.read().await;
        if handle_count != count {
            warn!(
                "[ProcessorPool] Stale count detected: handles={}, count={}, fixing",
@@ -117,6 +157,13 @@ impl ProcessorPool {
            let mut c = self.running_count.write().await;
            *c = handle_count;
        }
+        // 若 frame 或 time slot 超出 handle_count，降回合理值
+        if frame_count + time_count > handle_count {
+            let mut fc = self.running_frame_count.write().await;
+            let mut tc = self.running_time_count.write().await;
+            *fc = (*fc).min(handle_count);
+            *tc = (*tc).min(handle_count.saturating_sub(*fc));
+        }

        if handle_count == 0 && count == 0 {
            if let Err(e) = self
@@ -162,6 +209,7 @@ impl ProcessorPool {
        let job_id = task.job.id;
        let processor_type = task.processor_type;

+        let pipeline = task.processor_type.pipeline();
        let current_limit = self.current_max().await;
        {
            let mut count = self.running_count.write().await;
@@ -173,9 +221,17 @@ impl ProcessorPool {
            }
            *count += 1;
        }
+        // 遞增產線專屬 slot
+        match pipeline {
+            PipelineType::Frame => *self.running_frame_count.write().await += 1,
+            PipelineType::Time => *self.running_time_count.write().await += 1,
+            PipelineType::Cross => {} // cross pipeline uses global slot
+        }

        let running = self.running.clone();
        let running_count = self.running_count.clone();
+        let running_frame_count = self.running_frame_count.clone();
+        let running_time_count = self.running_time_count.clone();
        let child_pid: Arc<RwLock<Option<i32>>> = Arc::new(RwLock::new(None));
        running.write().await.insert(
            job_id,
@@ -205,6 +261,9 @@ impl ProcessorPool {
                job_id,
                running: running.clone(),
                running_count: running_count.clone(),
+                frame_count: running_frame_count.clone(),
+                time_count: running_time_count.clone(),
+                pipeline,
            };

            info!("Starting processor {} for job {}", processor_name, job.uuid);