Please Rotate Device

This site is optimized for portrait mode on mobile devices.

Über Mich

Ich stehe kurz vor dem Abschluss meines B.Sc. in Data Science mit starkem Fokus auf edge AI im MOT Bereich. Mich interessiert vor allem, wie Perception-Modelle in der realen Welt performen, wie etwa auf meinen selbstgebauten Drohnen, und nicht nur auf isolierten Benchmarks. Die folgenden Projekte zeigen, wie ich Custom-Hardware und robuste Software kombiniere, um End-to-End-Systeme für echte operative Herausforderungen zu entwickeln.

BURA Langstrecken Aufklärung

Eine vollständig 3D-gedruckte Drohne für Langstrecken- und mehrstündige Aufklärungsmissionen.


Hierarchisches Multi Object Tracking

Ein post-hoc hierarchisches Framework für YOLO und ähnliche Modelle, das unsichere Detektionen validiert und den Tracking-Recall drastisch verbessert.

Anki Automation

Eine Desktop-Anwendung zur KI-gestützten Erstellung von Anki-Lernkarten. Entwickelt für effizientes Lernen an der Universität. Jetzt als eigenständige Desktop-App mit geführtem Einrichtungsablauf und überarbeiteter Benutzeroberfläche verfügbar.

FRONT

Explain the Scaled Dot-Product Attention mechanism used in Transformers.

⋯ MORE CONTENT
BACK

The attention mechanism computes a weighted sum of values based on the compatibility of the query with corresponding keys:

\[ \text{Attention}(Q, K, V) = \text{softmax}\left(\frac{QK^T}{\sqrt{d_k}}\right)V \]

  • Q: Query matrix
  • K: Key matrix
  • V: Value matrix
  • \( d_k \): Dimension of keys (scaling factor to prevent vanishing gradients in softmax)
⋯ MORE CONTENT
FRONT

What is the objective function optimized in a Variational Autoencoder (VAE)?

⋯ MORE CONTENT
BACK

VAEs maximize the Evidence Lower Bound (ELBO):

\[ \mathcal{L}(\theta, \phi; x) = \mathbb{E}_{q_\phi(z|x)}[\log p_\theta(x|z)] - D_{KL}(q_\phi(z|x) || p(z)) \]

Components:
  • First term: Reconstruction Loss (how well we decode)
  • Second term: KL Divergence (regularization towards prior \( p(z) \))
⋯ MORE CONTENT
FRONT

Implement a Python decorator @time_execution that measures function runtime.

⋯ MORE CONTENT
BACK
import time
import functools

def time_execution(func):
    @functools.wraps(func)
    def wrapper(*args, **kwargs):
        start = time.perf_counter()
        result = func(*args, **kwargs)
        end = time.perf_counter()
        print(f"{func.__name__} took {end - start:.4f}s")
        return result
    return wrapper

@time_execution
def heavy_compute():
    # ... implementation
    pass
⋯ MORE CONTENT
FRONT

Describe the update rules for the Adam optimizer.

⋯ MORE CONTENT
BACK

Adam (Adaptive Moment Estimation) maintains moving averages of gradients and squared gradients:

\[ m_t = \beta_1 m_{t-1} + (1-\beta_1)g_t \]
\[ v_t = \beta_2 v_{t-1} + (1-\beta_2)g_t^2 \]

Bias-corrected estimates:

\[ \hat{m}_t = \frac{m_t}{1-\beta_1^t}, \quad \hat{v}_t = \frac{v_t}{1-\beta_2^t} \]

Parameter update:

\[ \theta_{t+1} = \theta_t - \frac{\eta}{\sqrt{\hat{v}_t} + \epsilon} \hat{m}_t \]

⋯ MORE CONTENT

Synthetic Data Generator

Ein modulares C++ Unreal Engine 5.7 Plugin zur automatisierten Erzeugung fotorealistischer Trainingsdaten inklusive Instanzmasken, Tiefe und Metadaten, reproduzierbar in interaktiven und Headless-Batches.

Kontakt

Ich freue mich über neue Projekte, Kooperationen oder Austausch.
Schreiben Sie mir gerne!