Card: Pooja's review (numbered citations, intro order, stairs caption, PerceptronAI author)

README.md

@@ -28,12 +28,12 @@ configs:
 
 This card follows the [EgoHome blog post](https://www.perceptron.inc/blog/egohome), which also has the full-clip videos for each annotation type.
 
+As part of our announcement of [Perceptron Egocentric 1.0](https://www.perceptron.inc/blog/introducing-perceptron-egocentric-1-0), we are releasing and open-sourcing EgoHome, a new benchmark containing 2 hours of human-annotated egocentric video of people performing various household tasks.
+
 ![The four annotation types on the same 5 s of one clip: hand boxes, subtasks, hand actions and hand captions](assets/egohome_header_2x2.gif)
 
 **Figure 1.** All four of our annotation types on the same five seconds of one clip.
 
-As part of our announcement of [Perceptron Egocentric 1.0](https://www.perceptron.inc/blog/introducing-perceptron-egocentric-1-0), we are releasing and open-sourcing EgoHome, a new benchmark containing 2 hours of human-annotated egocentric video of people performing various household tasks.
-
 The annotations span four different categories:
 
 - Hand location bounding boxes
@@ -45,17 +45,17 @@ We release the raw annotations as well as code for evaluation metrics, which we
 
 ### A new standard for Egocentric annotations
 
-Rich egocentric annotation is of increasing interest in the physical AI and robotics world, largely because of recent work demonstrating that it is highly effective as a scalable pretraining objective for world action models and vision language action models for [robotic manipulation tasks](https://www.dyna.co/dyna-2). More broadly, recent work has established both the volume and the diversity of pretraining data as essential for reliable control: Physical Intelligence has shown that [compositional generalization](https://www.pi.website/blog/pi07) can emerge from diverse multimodal data and conditioning, Generalist that [task-specific mastery](https://generalistai.com/blog/gen-1) can be achieved by scaling high-fidelity physical-interaction data, and Dyna that [scaling egocentric data](https://www.dyna.co/dyna-2) is one viable path to human-level control.
+Rich egocentric annotation is of increasing interest in the physical AI and robotics world, largely because of recent work demonstrating that it is highly effective as a scalable pretraining objective for world action models and vision language action models for [robotic manipulation tasks](https://www.dyna.co/dyna-2) [[1](#references)]. More broadly, recent work has established both the volume and the diversity of pretraining data as essential for reliable control: Physical Intelligence has shown that [compositional generalization](https://www.pi.website/blog/pi07) can emerge from diverse multimodal data and conditioning [[2](#references)], Generalist that [task-specific mastery](https://generalistai.com/blog/gen-1) can be achieved by scaling high-fidelity physical-interaction data [[3](#references)], and Dyna that [scaling egocentric data](https://www.dyna.co/dyna-2) is one viable path to human-level control [[1](#references)].
 
 Despite this increasing attention on egocentric data, it is still not perfectly well-understood which set of annotations serve as the best pretraining objectives or how best to measure the quality of synthetic labeling harnesses for these tasks.
 
-The most widely adopted egocentric benchmark for subtask segmentation, [WGO-Bench](https://macrodata.co/blog/annotating-robot-video-subtasks), has been a remarkably useful contribution to the ecosystem and has since become a de facto standard for evaluating the quality of a model or harness on describing the events in an egocentric video. Despite this, it still has several gaps in coverage which we seek to address with this release.
+The most widely adopted egocentric benchmark for subtask segmentation, [WGO-Bench](https://macrodata.co/blog/annotating-robot-video-subtasks) [[4](#references)], has been a remarkably useful contribution to the ecosystem and has since become a de facto standard for evaluating the quality of a model or harness on describing the events in an egocentric video. Despite this, it still has several gaps in coverage which we seek to address with this release.
 
 **Perceptron’s EgoHome** benchmark addresses a few gaps left behind by the current state of egocentric benchmarking:
 
 - We introduce benchmarking for four types of annotation all on the same video sources: hand box location, subtask timelines, hand action classification (from the Ego4D closed vocab), and hand action freeform captions.
 - We provide a large number of annotations at the sub-second scale: 3,371 segments under 1 s and 773 under 0.5 s.
-- Finally, we introduce a new flexible metric we call permissible subsegment Jaccard. An LLM-as-judge based technique which accommodates the fact that there are many permissible ways to break down the granularity of a task.
+- Finally, we introduce a new flexible metric we call permissible subsegment Jaccard: an LLM-as-judge based technique which accommodates the fact that there are many permissible ways to break down the granularity of a task.
 
 ## Benchmark overview
 
@@ -89,7 +89,7 @@ We annotate every hand box with human annotators at 10 fps then apply proprietar
 
 #### Hand actions
 
-We have annotators choose an accurate hand action label from the closed vocabulary [Ego4D action schema](https://ego4d-data.org/docs/benchmarks/hands-and-objects/); there are 118 total actions to choose from (47 occur in the release).
+We have annotators choose an accurate hand action label from the closed vocabulary [Ego4D action schema](https://ego4d-data.org/docs/benchmarks/hands-and-objects/) [[5](#references)]; there are 118 total actions to choose from (47 occur in the release).
 
 #### Hand captions
 
@@ -132,9 +132,9 @@ The table below demonstrates how various predictions would be graded under our m
 | C. “sweeping the stairs” 0–3 s, “putting the broom away” 3–5 s | first piece agrees: (3 + 3) / (5 + 5) = **0.6** | IoU 0.6, below 0.75 → **0.0** |
 | D. “carrying the broom” 0–5 s | judge says it’s a different action → **0.0** | timing matches but the label doesn’t → **0.0** |
 
-**Table 2.** Four model outputs for one human segment, “sweeping the stairs” from 0 to 5 s, scored by permissible subsegment Jaccard and WGO-Bench semantic F1. Solid segments agree with the human label under the judge; dashed segments do not.
+**Table 2.** Four model outputs for one human segment, “sweeping the stairs” from 0 to 5 s, scored by permissible subsegment Jaccard and WGO-Bench semantic F1. The ground-truth track is the human label; the model track is the prediction. Solid segments agree with the ground truth under the judge; dashed segments do not.
 
-While calibrating annotators, we frequently ran into Case B: a valid, more detailed breakdown gets a score of zero under one-to-one matching and full credit under ours. Case C shows we still give partial credit for the part the model got right, where a fixed IoU cutoff gives nothing. (WGO-Bench’s semantic F1 pairs segments one-to-one, counts a pair only when its temporal IoU is at least 0.75 and the judge accepts the labels).
+While calibrating annotators, we frequently ran into Case B, as illustrated in Table 2: a valid, more detailed breakdown gets a score of zero under one-to-one matching and full credit under ours. Case C shows we still give partial credit for the part the model got right, where a fixed IoU cutoff gives nothing. (WGO-Bench’s semantic F1 pairs segments one-to-one, counts a pair only when its temporal IoU is at least 0.75 and the judge accepts the labels).
 
 We found that this metric adhered much more closely to subjective measures of fairness when evaluating human annotators’ performance during calibration, thus we opted to release it as our official metric for the benchmark.
 
@@ -287,7 +287,7 @@ There is one row per clip. Times are seconds from the clip start. Boxes are `[x_
 Mathematically this can be expressed as
 
 $$
-\operatorname{CHF} = \frac{\displaystyle\sum_{(f,h) \in G} \mathbb{1}\left[\operatorname{Centroid}(\hat{p}_{f,h}) \in g_{f,h}\right]}{|G| + |P_{\text{out}}|}
+\mathrm{CHF} = \frac{\displaystyle\sum_{(f,h) \in G} \mathbb{1}\left[\mathrm{Centroid}(\hat{p}_{f,h}) \in g_{f,h}\right]}{|G| + |P_{\text{out}}|}
 $$
 
 Where G is the set of (frame, hand) pairs with a human box at 30 fps and g(f,h) is that box. p̂(f,h) is the model’s box for hand h at frame f: its own box at that frame, or the linear interpolation between its two nearest boxes; frames outside the span the model covered count as misses. P_out is the set of predicted boxes outside the frames the human labels cover, and the indicator is 1 when the centroid of p̂(f,h) falls inside g(f,h).
@@ -297,7 +297,7 @@ Where G is the set of (frame, hand) pairs with a human box at 30 fps and g(f,h)
 It can be expressed mathematically as follows:
 
 $$
-\operatorname{PSJ}(P, T) = \frac{\displaystyle\sum_{p \in P} m(p, T) + \sum_{t \in T} m(t, P)}{\displaystyle\sum_{p \in P} |I(p)| + \sum_{t \in T} |I(t)|}
+\mathrm{PSJ}(P, T) = \frac{\displaystyle\sum_{p \in P} m(p, T) + \sum_{t \in T} m(t, P)}{\displaystyle\sum_{p \in P} |I(p)| + \sum_{t \in T} |I(t)|}
 $$
 
 $$
@@ -447,7 +447,7 @@ Here is the percentage of time the right and left hand are on screen in the corp
 ```bibtex
 @misc{perceptron2026egohome,
   title={Perceptron {EgoHome}: A new benchmark for dense, sub-second egocentric understanding},
-  author={{Perceptron}},
+  author={{PerceptronAI}},
   year={2026},
   month={nov},
   url={https://www.perceptron.inc/blog/egohome}