Eve library

ISIS v2 Vehicle Interior Pack Control-Conditioned Workflow Graph

Guide

Revision 1

On this page

Generated: 2026-03-01T10:38:36.755Z Checklist: MXA-0274 (ISIS2-00472) Asset Family: vehicle-interior-pack Phase: Phase 4 | Milestone: M3 Launch (2026-07-27 to 2026-10-30)

Objective#

Define a control-conditioned workflow graph for vehicle-interior-pack that supports pose/depth/segmentation/canny constraints while preserving deterministic interior-generation defaults.

Deterministic Defaults#

Setting Value
checkpointModel sd_xl_base_1.0.safetensors
seed 2026120103
width 2048
height 1152
steps 36
cfgScale 6
samplerName dpmpp_2m_sde
scheduler karras
denoise 1
poseControlModel control_v11p_sd15_openpose.safetensors
depthControlModel control_v11f1p_sd15_depth.safetensors
segmentationControlModel control_v11p_sd15_seg.safetensors
cannyControlModel control_v11p_sd15_canny.safetensors
poseStrength 0.66
depthStrength 0.72
segmentationStrength 0.6
cannyStrength 0.65
batchSize 1
outputFormat png
outputChannels 4

Input Contract#

Input Type Required Default
pose_map image no n/a
depth_map image no n/a
segmentation_map image no n/a
canny_map image no n/a
use_pose_control boolean yes true
use_depth_control boolean yes true
use_segmentation_control boolean yes false
use_canny_control boolean yes true
pose_strength float yes 0.66
depth_strength float yes 0.72
segmentation_strength float yes 0.6
canny_strength float yes 0.65
cabin_type enum yes sports-coupe
seating_configuration enum yes 2-plus-2
dashboard_layout enum yes driver-centric-cockpit
steering_side enum yes left-hand-drive
upholstery_material enum yes nappa-leather
trim_material enum yes brushed-aluminum
ambient_lighting_profile enum yes subtle-warm
instrument_cluster_style enum yes full-digital
infotainment_layout enum yes single-center-screen
center_console_configuration enum yes floating-console
climate_control_mode enum yes tri-zone
interior_accessory_package enum yes performance
window_tint_profile enum yes light-smoke
environment_context enum yes studio-cabin-rig
camera_view_set enum yes driver-seat-hero
lighting_setup enum yes neutral-studio
weather_condition enum yes dry
color_palette_profile string yes charcoal leather, satin aluminum, matte black controls, warm ambient strip, cool display glow
style_preset enum yes photoreal-catalog
negative_prompt string no blurry, low resolution, warped dashboard geometry, floating controls, stretched seat geometry, duplicated steering wheel, text, watermark, jpeg artifacts
seed integer yes 2026120103
width integer yes 2048
height integer yes 1152
steps integer yes 36
cfg_scale float yes 6

Graph Nodes#

Node ID Class Role Output Count
n1 LoadImage pose-control-input 2
n2 LoadImage depth-control-input 2
n3 LoadImage segmentation-control-input 2
n4 LoadImage canny-control-input 2
n5 CheckpointLoaderSimple model-load 3
n6 ControlNetLoader pose-control-load 1
n7 ControlNetLoader depth-control-load 1
n8 ControlNetLoader segmentation-control-load 1
n9 ControlNetLoader canny-control-load 1
n10 CLIPTextEncode positive-conditioning 1
n11 CLIPTextEncode negative-conditioning 1
n12 ControlNetApply pose-control-apply 1
n13 ControlNetApply depth-control-apply 1
n14 ControlNetApply segmentation-control-apply 1
n15 ControlNetApply canny-control-apply 1
n16 EmptyLatentImage latent-init 1
n17 KSampler latent-sample 1
n18 VAEDecode image-decode 1
n19 SaveImage artifact-output 0

Graph Edges#

Edge ID From To
e1 n5:0 n17.model
e2 n5:1 n10.clip
e3 n5:1 n11.clip
e4 n5:2 n18.vae
e5 n10:0 n12.conditioning
e6 n6:0 n12.control_net
e7 n1:0 n12.image
e8 n12:0 n13.conditioning
e9 n7:0 n13.control_net
e10 n2:0 n13.image
e11 n13:0 n14.conditioning
e12 n8:0 n14.control_net
e13 n3:0 n14.image
e14 n14:0 n15.conditioning
e15 n9:0 n15.control_net
e16 n4:0 n15.image
e17 n15:0 n17.positive
e18 n11:0 n17.negative
e19 n16:0 n17.latent_image
e20 n17:0 n18.samples
e21 n18:0 n19.images

Topological Node Order#

n1 -> n16 -> n2 -> n3 -> n4 -> n5 -> n10 -> n11 -> n6 -> n12 -> n7 -> n13 -> n8 -> n14 -> n9 -> n15 -> n17 -> n18 -> n19

Research Alignment#

Workflow Path Control Nodes
AnimateDiff with ControlNet libs/isis/workflows/src/workflows/video/animatediff-controlnet.json 3
3D Texture Generator libs/isis/workflows/src/workflows/texture/3d-texture.json 0
HDRI Environment Map Generator libs/isis/workflows/src/workflows/texture/environment-map.json 0

Acceptance Criteria#

  • Graph is acyclic and topologically sortable.
  • Pose/depth/segmentation/canny control paths are explicitly represented through ControlNet loaders and apply nodes.
  • At least one control mode can be enabled per run without breaking deterministic sampler/model defaults.
  • Output contract aligns to MXA-0271 required output classes.
  • Cabin layout, seating configuration, dashboard architecture, and trim/lighting controls are first-class required inputs for controlled generation.

Traceability#

  • Checklist item: MXA-0274
  • Ticket key: ISIS2-00472
  • Source references:
    • docs/domains/isis/remediation-v2/ISIS_REMEDIATION_V2_TICKETS.json
    • docs/domains/isis/remediation-v2/ISIS_V2_VEHICLE_INTERIOR_PACK_REQUIREMENTS_BRIEF.json
    • docs/domains/isis/remediation-v2/ISIS_V2_VEHICLE_INTERIOR_PACK_BASELINE_WORKFLOW_GRAPH.json
    • docs/domains/isis/remediation-v2/ISIS_V2_VEHICLE_INTERIOR_PACK_IMAGE_CONDITIONED_WORKFLOW_GRAPH.json
    • libs/isis/workflows/src/workflows/video/animatediff-controlnet.json
    • libs/isis/workflows/src/workflows/texture/3d-texture.json
    • libs/isis/workflows/src/workflows/texture/environment-map.json
    • docs/domains/isis/remediation-v2/ISIS_V2_CONTROL_TAXONOMY.json
    • docs/domains/isis/remediation-v2/ISIS_V2_OUTPUT_PACKAGE_TAXONOMY.json
    • ISIS_TODOS_2.md