Expand technical review with literature methodology and full architecture

docs/OrbitSight_Full_Technical_Proposal_Review.md

@@ -93,6 +93,86 @@ recall, F1, and AP/mAP at IoU 0.5. The competition additionally calls for end-to
 below 40 ms on an Intel Core i9-12900H CPU. There is no track-ID column, so tracking matters
 only when it improves the emitted detection stream.
 
+## Literature review and design lineage
+
+OrbitSight is an application-specific synthesis rather than a direct reproduction of one published
+network. Its design draws on neuromorphic sensing, recurrent event detection, anchor-free dense
+detection, model compression, and detection-assisted tracking.
+
+### Event cameras and space situational awareness
+
+Gallego et al. formalize the event-camera model: a pixel reports a change in log intensity rather
+than an absolute image sample. The literature emphasizes microsecond timing, high dynamic range,
+low motion blur, and sparse output, while warning that frame algorithms require an event
+representation or event-native computation. OrbitSight preserves polarity and within-window timing,
+but packs events into 40 ms tensors so convolutional and recurrent operations run efficiently on a
+CPU. [Gallego et al., *Event-based Vision: A Survey*, TPAMI 2020](https://doi.org/10.1109/TPAMI.2020.3008413).
+
+Afshar et al. demonstrated event-based RSO detection and tracking across sensors and observing sites.
+Ralph et al. subsequently studied astrometric calibration and source characterization for event-based
+space imaging. These works motivate OrbitSight's explicit sensor profiles, native-coordinate
+reconstruction, and treatment of preprocessing and coordinate mapping as part of the model contract.
+[Afshar et al., *Event-based Object Detection and Tracking for Space Situational Awareness*,
+2019](https://arxiv.org/abs/1911.08730); [Ralph et al., *Astrometric Calibration and Source
+Characterisation of Neuromorphic Event-based Cameras for Space Imaging*,
+2022](https://arxiv.org/abs/2211.09939).
+
+### Recurrent event-based object detection
+
+Perot et al.'s Recurrent Event-camera Detector showed that direct event representations combined with
+recurrent memory outperform feed-forward event detectors because state integrates evidence through
+time. RVT later showed that multi-stage recurrent backbones can obtain a favorable latency/accuracy
+tradeoff. OrbitSight adopts direct event tensors, multi-scale features, and persistent temporal state,
+but uses narrow convolutions and ConvLSTM instead of a transformer for an offline CPU target and tiny
+objects. [Perot et al., *Learning to Detect Objects with a 1 Megapixel Event Camera*, NeurIPS
+2020](https://proceedings.neurips.cc/paper/2020/hash/c213877427b46fa96cff6c39e837ccee-Abstract.html);
+[Gehrig and Scaramuzza, *Recurrent Vision Transformers for Object Detection With Event Cameras*,
+CVPR 2023](https://openaccess.thecvf.com/content/CVPR2023/html/Gehrig_Recurrent_Vision_Transformers_for_Object_Detection_With_Event_Cameras_CVPR_2023_paper.html).
+
+ConvLSTM replaces fully connected state transitions with spatial convolutions, retaining *where* weak
+evidence occurred. OrbitSight places one after every downsampling stage and uses dilations 1, 2, 4,
+and 8 to enlarge spatial context without changing tensor shapes. [Shi et al., *Convolutional LSTM
+Network*, NeurIPS 2015](https://proceedings.neurips.cc/paper/2015/hash/07563a3fe3bbe7e3ba84431ad9d055af-Abstract.html).
+
+### Anchor-free detection, imbalance, and localization
+
+FCOS established fully convolutional anchor-free prediction; YOLOX paired anchor-free boxes with a
+decoupled head. OrbitSight adopts those ideas with custom assignment and scoring for one RSO class.
+It predicts raw center offsets and log sizes on three scales without an anchor bank. [Tian et al.,
+*FCOS*, ICCV 2019](https://openaccess.thecvf.com/content_ICCV_2019/html/Tian_FCOS_Fully_Convolutional_One-Stage_Object_Detection_ICCV_2019_paper.html);
+[Ge et al., *YOLOX*, 2021](https://arxiv.org/abs/2107.08430).
+
+Dense event tensors create tens of thousands of background cells for at most one target. Focal loss
+suppresses easy negatives so hard examples control learning. OrbitSight uses sigmoid focal BCE for
+objectness. CIoU combines overlap, center distance, and aspect-ratio consistency, aligning training
+more closely with the IoU≥0.5 metric than independent coordinate L1. [Lin et al., *Focal Loss for
+Dense Object Detection*, ICCV 2017](https://openaccess.thecvf.com/content_ICCV_2017/papers/Lin_Focal_Loss_for_ICCV_2017_paper.pdf);
+[Zheng et al., *Distance-IoU Loss*, AAAI 2020](https://ojs.aaai.org/index.php/AAAI/article/view/6999).
+
+### Distillation and tracking
+
+Knowledge distillation transfers useful behavior from an expensive model or ensemble into a smaller
+deployable network. OrbitSight follows that deployment logic but does not blindly copy teacher output:
+human boxes remain authoritative, matched teacher boxes provide low-weight targets, and proposals in
+human-empty windows become negative evidence. [Hinton, Vinyals, and Dean, *Distilling the Knowledge
+in a Neural Network*, 2015](https://arxiv.org/abs/1503.02531).
+
+ByteTrack showed the value of associating lower-confidence detections. OrbitSight evaluated that
+two-tier idea, but production emits only stable global top-one at 0.10, making the nominal low tier
+unreachable. The final system is therefore a single-tier constant-velocity Kalman/coast tracker using
+the same association machinery. [Zhang et al., *ByteTrack*, ECCV
+2022](https://www.ecva.net/papers/eccv_2022/papers_ECCV/papers/136820001.pdf).
+
+### OrbitSight's design contribution
+
+- a polarity/recency representation with sensor-specific event denoising;
+- a full-resolution but ultra-narrow recurrent detector;
+- single-class objectness decoding without a redundant class multiplier;
+- selective, human-authoritative distillation from a sensor-routed teacher;
+- weights bound to preprocessing and coordinate-mapping fingerprints;
+- exact stable top-one decoding that avoids NMS without changing the selected box;
+- one evaluated detector/tracker/writer deployment contract.
+
 ## 2. Development history at a glance
 
 The following table separates comparable leakage-free results from historical development
@@ -406,53 +486,280 @@ checkpoint SHA-256 is:
 
 `5f176b9baffc3478b1aaada13277e2c6093ba02439b76b8330cb8d654f5f554d`
 
-## 10. Compact student architecture
+## 10. Final-model methodology
+
+This section describes how the final student was produced and evaluated. Section 11 then defines
+the frozen network and inference path component by component.
+
+### 10.1 Research questions and temporal unit
+
+Development asked whether one model could localize tiny RSOs across all three sensor families,
+inherit useful sensor-specific behavior from the heavy hybrid, and keep the complete deployed path
+below 40 ms. The atomic example is a half-open 40,000 µs event interval. Window phase is anchored to
+the sequence timestamp contract. Recurrent, BA, and tracker state are initialized at the start of a
+sequence, advanced exactly once per window, and destroyed at the boundary.
+
+Training uses ordered segments, not shuffled independent frames. Final distillation used 20 consecutive
+windows per segment—0.8 seconds of temporal context—and batches of 16 independent segments. Each batch
+member carries its own four hidden/cell state pairs before truncated backpropagation detaches them.
+
+### 10.2 Data preparation
+
+For every window the method:
+
+1. memory-maps and timestamp-slices events;
+2. applies the sensor-specific neighborhood/time BA rule;
+3. builds positive-count, negative-count, and latest-event-recency planes;
+4. area-resizes the planes to 640×640;
+5. maps human boxes with the same pixel-center convention;
+6. builds the optional per-event auxiliary target.
 
-### 10.1 Heavy teacher versus compact student
+Counts remain raw. Stateful BA, sensor FiLM, hot-pixel changes, and target-event thinning are disabled
+in the final route. Cache fingerprints bind the source, sensor profile, 40 ms phase, input size, count
+transform, area resize, and coordinate mapping. This prevents a shape-compatible checkpoint from being
+used with semantically different preprocessing, the failure that invalidated V6.
+
+### 10.3 Human-authoritative teacher evidence
+
+The V5/V7 router runs offline. Its pre-tracker proposals are stored in sidecars aligned one-to-one with
+student windows and marked as matched or unmatched against human GT. Post-tracker teacher boxes are
+diagnostic only. The selective objective is:
+
+```text
+L_total = L_human
+        + 0.15 L_teacher-objectness-on-matched-GT
+        + 0.25 L_teacher-box-on-matched-GT
+        + 0.35 L_teacher-empty-window-negative
+```
+
+Teacher scores below 0.05 are ignored. Matched proposals supply a BCE confidence target and Smooth-L1
+encoded-box target. A proposal in a human-empty window pushes the corresponding student objectness
+toward zero. An unmatched teacher proposal never creates a positive label. The teacher is therefore a
+ranking guide and hard-negative miner, not a replacement annotator.
+
+### 10.4 Optimization protocol
+
+The final A100 run used 36 epochs, 260 updates per epoch, and 9,360 optimizer steps:
+
+| Item | Final setting |
+|---|---:|
+| Optimizer | Adam |
+| Initial / minimum LR | 5×10⁻⁴ / 1×10⁻⁵ |
+| Warm-up / schedule | 100 steps / cosine decay |
+| Precision | FP32; TF32 disabled |
+| Gradient-norm clip | 1.0 |
+| Batch / segment | 16 / 20 windows |
+| Negative/positive target | 2.0 |
+| Maximum oversampling | 4.0× |
+| Human box loss | pixel-inclusive CIoU, weight 2.0 |
+| Auxiliary segmentation | BCE, weight 0.5 |
+| Objectness target floor | 1.0 throughout |
+| Quality head / loss | disabled / 0.0 |
+| Seed | 1337 |
+
+Adam state, scheduler, RNG, batching order, partial epoch progress, preprocessing fingerprint, and the
+training contract were included in exact-resume checkpoints. A remote disconnect therefore resumed
+the same stochastic run rather than launching a similar but different experiment.
+
+### 10.5 Evaluation methodology
+
+Predictions are scored only after native mapping, calibration, tracking, clipping, top-one selection,
+sensor output gating, and integer writer projection. IoU uses the official pixel-inclusive convention.
+Metrics are reported for the three-sequence leakage-free aggregate and all four local recordings.
+Because every remaining recording has influenced diagnostics, these are reproducibility and local
+comparison results—not blind-generalization proof. Latency is measured after warm-up at streaming
+batch one and reports mean, per-sequence p95, and maximum rather than forward-pass mean alone.
+
+## 11. Final model architecture: complete component reference
+
+### 11.1 Frozen architecture contract
+
+| Property | Final value |
+|---|---:|
+| Class | `OrbitSightDetector` |
+| Inference input | 1×3×640×640 |
+| Backbone widths | 4, 8, 16, 32 |
+| ConvLSTM dilations | 1, 2, 4, 8 |
+| Detection scales | stride 4, 8, 16 |
+| Head width | 16 |
+| Classes | one RSO foreground class |
+| Stored parameters | 135,111 |
+| Quality head / sensor FiLM | absent / absent |
+| Auxiliary segmentation | stored; training-only execution |
+| Classification towers | stored for compatibility; never executed |
+
+![Exact final student architecture](full_proposal_assets/final_student_architecture.png)
+
+### 11.2 Compression decision
 
 | Component | Heavy V5/V7 branch | CPU student |
 |---|---:|---:|
 | Input | 3×640×640 | 3×640×640 |
 | Backbone widths | 32, 64, 128, 256 | 4, 8, 16, 32 |
-| Detection-head width | 128 | 16 |
-| Recurrent dilations | 1, 2, 4, 8 | 1, 2, 4, 8 |
+| Head width | 128 | 16 |
+| Dilations | 1, 2, 4, 8 | 1, 2, 4, 8 |
 | Parameters | 8,588,339 | 135,111 |
-| Relative size | 1× | 63.6× smaller |
-| Runtime role | Training teacher | Deployed CPU model |
+| Runtime role | Offline teacher | Deployed model |
 
 ![Heavy teacher versus compact student](full_proposal_assets/heavy_vs_student.png)
 
-The student does not shrink every dimension. It preserves the 640×640 input, stride-4 output,
-four recurrent stages, and multi-scale detection pattern. It primarily removes channel width.
-That is the correct trade for tiny targets: channel width costs computation, while spatial
-resolution protects localization.
+The 63.6× reduction comes mainly from channel width. Input resolution, recurrence, the finest stride-4
+grid, and three detection scales remain because spatial quantization is expensive for 6–11 pixel targets.
 
-### 10.2 What happens inside one window
+### 11.3 Input representation
 
-1. **Input tensor:** positive count, negative count, and recency arrive as a 640×640 tensor.
-2. **Stage 1:** a narrow convolution extracts edges and local event clusters at high spatial
-   resolution. A ConvLSTM mixes this evidence with the previous window.
-3. **Stages 2–4:** spatial resolution decreases while channels and context grow. Dilations
-   allow later stages to see wider neighborhoods without a large kernel.
-4. **Feature pyramid:** features at different scales are passed to the detector so a tiny,
-   medium, or bright elongated response can be localized at an appropriate level.
-5. **Anchor-free head:** each grid location predicts objectness, center offset, width, and
-   height. The student is single-class: the question is RSO versus background.
-6. **NMS and top-one ranking:** overlapping proposals are suppressed and the best remaining
-   proposal is sent to tracking.
+`X ∈ R^(B×3×640×640)` is float32. Channel 0 counts positive-polarity events; channel 1 counts negative
+events; channel 2 records normalized recency of the latest event. Empty pixels are zero. This preserves
+brightness-change sign and within-window order without reconstructing an intensity image.
 
-An auxiliary segmentation head helps shape features during training but is skipped at
-inference. The quality head and sensor-conditioning branch are disabled. Keeping unused
-training-only work out of inference protects the latency budget.
+### 11.4 Stem
 
-### 10.3 Why recurrence remains in such a small network
+```text
+(B,3,640,640)
+ → Conv2d(3→4, kernel 3, stride 1, pad 1)
+ → BatchNorm2d(4) → SiLU
+ = (B,4,640,640)
+```
+
+The stem has 120 parameters. SiLU is `x·sigmoid(x)`. There is no pooling or tokenization.
+
+### 11.5 Four downsampling and recurrent stages
+
+Each stage performs 3×3 stride-2 convolution, BatchNorm, SiLU, then ConvLSTM. There are no residual
+blocks, attention layers, transformer tokens, or active sensor-conditioning layers.
+
+| Stage | Down mapping | Feature and state | Dilation | Down params | ConvLSTM params | Head input? |
+|---|---|---|---:|---:|---:|---|
+| 1 | 4→4 | B×4×320×320 | 1 | 156 | 1,168 | No |
+| 2 | 4→8 | B×8×160×160 | 2 | 312 | 4,640 | Yes, stride 4 |
+| 3 | 8→16 | B×16×80×80 | 4 | 1,200 | 18,496 | Yes, stride 8 |
+| 4 | 16→32 | B×32×40×40 | 8 | 4,704 | 73,856 | Yes, stride 16 |
+
+Stage 1 still matters: its high-resolution recurrent output feeds all later stages. The last three
+tensors are independent multi-scale features. Unlike a classical FPN, there is **no top-down pathway,
+lateral addition, or cross-scale fusion**.
+
+### 11.6 Every ConvLSTM operation
 
-On low-SNR windows, the RSO may contribute too few events for a reliable single-window
-decision. ConvLSTM state acts like a learned short-term memory: it can retain a weak pattern
-that appears consistently over consecutive windows. Removing recurrence would save compute,
-but it would discard the most direct answer to signal starvation.
+At time `t`, current feature `X_t` and previous hidden state `H_(t−1)` are concatenated. One dilated
+3×3 convolution emits four channel groups:
 
-## 11. Detection to tracking
+```text
+[I,F,O,G] = Conv3×3_dilated([X_t,H_(t−1)])
+i=sigmoid(I), f=sigmoid(F), o=sigmoid(O), g=tanh(G)
+C_t = f ⊙ C_(t−1) + i ⊙ g
+H_t = o ⊙ tanh(C_t)
+```
+
+The input, forget, and output gates govern writing, retaining, and exposing memory. Both `H_t` and
+`C_t` persist. At inference the four pairs have shapes 1×4×320×320, 1×8×160×160,
+1×16×80×80, and 1×32×40×40. Dilation gives effective 3×3, 5×5, 9×9, and 17×17 footprints without
+adding weights. State begins at zero and resets only at sequence boundaries.
+
+### 11.7 Three active detection heads
+
+For each returned feature, the active regression/objectness tower is:
+
+```text
+feature C×H×W
+ → Conv2d(C→16,3×3,pad1) → SiLU
+ → Conv2d(16→16,3×3,pad1) → SiLU
+ ├→ Conv2d(16→4,1×1): raw box vector
+ └→ Conv2d(16→1,1×1): objectness logit
+```
+
+| Scale | Feature | Tower + box + objectness params | Outputs |
+|---|---:|---:|---|
+| stride 4 | 8×160×160 | 3,488 + 68 + 17 | 4×160×160 and 1×160×160 |
+| stride 8 | 16×80×80 | 4,640 + 68 + 17 | 4×80×80 and 1×80×80 |
+| stride 16 | 32×40×40 | 6,944 + 68 + 17 | 4×40×40 and 1×40×40 |
+
+There are 25,600 + 6,400 + 1,600 = **33,600 candidate cells**. Four regression channels encode raw
+x/y offsets and log width/height. Objectness alone asks whether an RSO occupies the cell.
+
+### 11.8 Compatibility classification modules
+
+Three classification towers and 1×1 predictors remain in the checkpoint so historical state-dictionary
+keys load strictly. They contain 15,123 parameters, but `compute_cls=False`: their convolutions are not
+executed, their logits are not trained by the final loss, and the decoder never reads them. In a
+single-class problem, multiplying objectness by a redundant class probability would square/compress
+confidence without adding information. Final score is `sigmoid(objectness)`. No quality head exists.
+
+### 11.9 Auxiliary segmentation head
+
+A nine-parameter 1×1 convolution maps the eight-channel stride-4 feature to one 160×160 logit plane.
+Per-event labels supervise it with BCE at weight 0.5, helping the finest features retain target-local
+structure. It runs only while `model.training` is true; deployment skips it entirely.
+
+### 11.10 Target assignment and hard loss
+
+A human box is assigned to the scale whose `4×stride` nominal size is closest to its maximum dimension.
+The rounded center and radius-1 neighborhood—up to 3×3 cells—are positive. At cell `(g_x,g_y)`:
+
+```text
+t_x=c_x/s−g_x,  t_y=c_y/s−g_y,
+t_w=w/s,        t_h=h/s
+```
+
+Objectness uses focal BCE with alpha 0.25 and gamma 2.0, normalized by positive assigned cells and
+clamped to one on empty windows. The positive objectness floor was 1.0 during final distillation.
+Regression uses pixel-inclusive CIoU: `1−IoU`, normalized center distance, and aspect-ratio penalty.
+It therefore supplies gradient even before two tiny boxes overlap. Auxiliary BCE and the selective KD
+terms described in Section 10 complete the objective.
+
+### 11.11 Anchor-free decode
+
+For raw `(r_x,r_y,r_w,r_h)` at grid cell `(g_x,g_y)` and stride `s`:
+
+```text
+c_x=(g_x+r_x)s       c_y=(g_y+r_y)s
+w=exp(clamp(r_w,6))s h=exp(clamp(r_h,6))s
+score=sigmoid(objectness)
+```
+
+Offsets are intentionally raw, matching training. Exponentiation guarantees positive sizes and the
+upper clamp prevents overflow.
+
+### 11.12 Exact stable top-one decode
+
+Production requests one proposal. The decoder concatenates three score maps in fixed scale/grid order,
+masks values below 0.10, and applies `argmax`. Exact ties select the first index deterministically. Only
+the winning cell's box is decoded.
+
+This exactly matches the first result of score-ordered NMS: the global maximum is selected before any
+overlap comparison can suppress it. Thus production avoids materializing/sorting 33,600 boxes and the
+Python NMS loop without approximating or changing the selected prediction. Generic pixel-inclusive NMS
+and its 1,000-candidate pre-bound remain only for multi-detection diagnostics.
+
+### 11.13 Native-coordinate reconstruction
+
+The winning box is inverted through recorded area-resize geometry with pixel-center rounding. DVX
+width/height receives the frozen 0.98 calibration before association. Geometry is clipped in native
+sensor coordinates, but remains floating point until the writer performs final legal integer projection.
+
+### 11.14 Complete parameter inventory
+
+| Module group | Parameters | Deployment execution |
+|---|---:|---|
+| Stem | 120 | Yes |
+| Four downsampling blocks | 6,372 | Yes |
+| Four ConvLSTM cells | 98,160 | Yes |
+| Three regression towers | 15,072 | Yes |
+| Box predictors | 204 | Yes |
+| Objectness predictors | 51 | Yes |
+| Compatibility classification modules | 15,123 | No |
+| Auxiliary segmentation | 9 | No |
+| **Total** | **135,111** | — |
+
+### 11.15 End-to-end state machine
+
+For each sequence: identify sensor and reset state; slice 40 ms; denoise; form the 3×640×640 tensor;
+run stem and four ConvLSTMs; run three active heads; choose/decode global top-one; map/calibrate native
+geometry; update Kalman/coast state; retain at most one box; apply the sensor gate; project integer
+fields; write the row; and reset everything before another sequence. Docker latency covers this path,
+not merely the neural forward call.
+
+## 12. Detection to tracking
 
 The detector proposes appearance-based boxes. The tracker imposes short-horizon physical
 consistency. Each track maintains:
@@ -478,7 +785,7 @@ This is intentionally simple. More elaborate motion gates were tested, but measu
 motion was slow enough that consecutive boxes overlap. IoU association was therefore more
 reliable than a complex motion-only discriminator.
 
-## 12. Distilled student results
+## 13. Distilled student results
 
 Before the final no-retrain adjustments, the distilled checkpoint produced:
 
@@ -503,7 +810,7 @@ The last number is why both percentiles and maxima are reported. The system comf
 the steady-state p95 objective locally, but scheduling outliers still existed in the longer
 profile.
 
-## 13. Final no-retrain improvements
+## 14. Final no-retrain improvements
 
 With the checkpoint and architecture frozen, only low-cost constants were screened:
 
@@ -534,7 +841,7 @@ policy remained comfortably under the local p95 limit and introduced no expensiv
 The gains are modest but consistent across both aggregate views. No weight, layer, or tensor
 shape changed.
 
-## 14. Current per-sequence behavior
+## 15. Current per-sequence behavior
 
 | Recording | Sensor | F1 | AP | Precision | Recall | Main interpretation |
 |---|---|---:|---:|---:|---:|---|
@@ -554,7 +861,7 @@ one is primarily ranking/clutter, and the other is primarily missing evidence.
 
 ![Representative localization failures](../competition_ready_improved_cpu_student/visualizations/promoted_failure_cases.png)
 
-## 15. Docker and reproducibility architecture
+## 16. Docker and reproducibility architecture
 
 The final Linux/amd64 image is designed to be inspected and run without network access. It
 contains:
@@ -587,7 +894,7 @@ Archive validation checks the Linux/amd64 manifest, layer diff IDs, automatic co
 mount contract, output schema, model hash, runtime dependencies, calibration constants,
 tracker settings, and baked model-structure text.
 
-## 16. Evidence boundaries and scientific honesty
+## 17. Evidence boundaries and scientific honesty
 
 Three different claims must not be confused:
 
@@ -605,7 +912,7 @@ The leakage-free aggregate removes the known DAVIS session overlap, but the rema
 have still been used for diagnostics. “Leakage-free” here means free of that known training-
 session contamination, not untouched by all development decisions.
 
-## 17. Why the final student is the recommended submission
+## 18. Why the final student is the recommended submission
 
 The heavy hybrid has the best local accuracy but fails the CPU requirement by an order of
 magnitude. V5 has a cleaner scientific history but its corrected full-tail latency exceeded
@@ -625,7 +932,7 @@ student offers:
 This is a practical systems decision rather than a claim that the student is universally
 more accurate than every predecessor.
 
-## 18. Recommended next work
+## 19. Recommended next work
 
 If new recordings become available, the next work should be data-centered rather than another
 threshold sweep:
@@ -645,7 +952,7 @@ For model research, the highest-value directions are:
 - measure container startup, file reading, formatting, workbook creation, and disk output on
   the organizer's exact i9-12900H hardware.
 
-## 19. Conclusion
+## 20. Conclusion
 
 OrbitSight progressed from an honest sub-0.1 baseline to a complete real-time CPU system by
 treating data contracts, localization, temporal reasoning, deployment, and evaluation as one
@@ -661,14 +968,28 @@ transparent evidence, explicit limitations, and a reproducible offline delivery
 
 ## Selected references
 
-1. Y. Alkendi et al., “Dynamic Space Object Detection with Neuromorphic Vision Sensors,”
+1. G. Gallego et al., “Event-based Vision: A Survey,” IEEE TPAMI, 2020.
+   DOI: 10.1109/TPAMI.2020.3008413.
+2. S. Afshar et al., “Event-based Object Detection and Tracking for Space Situational
+   Awareness,” arXiv:1911.08730, 2019.
+3. N. O. Ralph et al., “Astrometric Calibration and Source Characterisation of the Latest
+   Generation Neuromorphic Event-based Cameras for Space Imaging,” arXiv:2211.09939, 2022.
+4. Y. Alkendi et al., “Dynamic Space Object Detection with Neuromorphic Vision Sensors,”
    European Conference on Space Debris, 2025.
-2. M. Gehrig and D. Scaramuzza, “Recurrent Vision Transformers for Object Detection with
+5. E. Perot et al., “Learning to Detect Objects with a 1 Megapixel Event Camera,” NeurIPS, 2020.
+6. M. Gehrig and D. Scaramuzza, “Recurrent Vision Transformers for Object Detection with
    Event Cameras,” CVPR, 2023.
-3. Y. Zhang et al., “ByteTrack: Multi-Object Tracking by Associating Every Detection Box,”
-   ECCV, 2022.
-4. Z. Zheng et al., “Distance-IoU Loss: Faster and Better Learning for Bounding Box
-   Regression,” AAAI, 2020.
+7. X. Shi et al., “Convolutional LSTM Network: A Machine Learning Approach for Precipitation
+   Nowcasting,” NeurIPS, 2015.
+8. Z. Tian et al., “FCOS: Fully Convolutional One-Stage Object Detection,” ICCV, 2019.
+9. Z. Ge et al., “YOLOX: Exceeding YOLO Series in 2021,” arXiv:2107.08430, 2021.
+10. T.-Y. Lin et al., “Focal Loss for Dense Object Detection,” ICCV, 2017.
+11. Z. Zheng et al., “Distance-IoU Loss: Faster and Better Learning for Bounding Box
+    Regression,” AAAI, 2020. DOI: 10.1609/aaai.v34i07.6999.
+12. G. Hinton, O. Vinyals, and J. Dean, “Distilling the Knowledge in a Neural Network,”
+    arXiv:1503.02531, 2015.
+13. Y. Zhang et al., “ByteTrack: Multi-Object Tracking by Associating Every Detection Box,”
+    ECCV, 2022.
 
 ## Primary local evidence