<h1 id="v2-version-is-out">V2 version is out</h1>Trained using a larger and more diverse dataset.Find training details on my HF page: <a target="_blank" rel="ugc" href="https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-Cameraman_v2">https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-Cameraman_v2</a>Example workflow is here: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-cameraman-v2/README.md">https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-cameraman-v2/README.md</a> It works with both Text-to-Video and Image-to-Video. Tips<ul><li>Resolution: based on my testing, the higher the resolution, the more closely the reference camera motion is followed. I wouldn't go below 960x512 for the first pass.</li><li>Image (conditioning) strength: use an image strength of 0.5 or 0.7 for more motion.</li><li>The prompt matters a lot — it strongly affects the camera movement. If the output doesn't follow the reference camera motion, you can try:<ul><li>leaving the prompt empty (in some cases this works best),</li><li>a different seed,</li><li>describing the camera motion explicitly, at least at a high level.</li></ul></li></ul><h1 id="v1">V1</h1>Let me share with you this IC-LoRA I just cooked which is intended to replicate camera movements from reference videos.You can find more training details and also can download the lora from here: <a target="_blank" rel="ugc" href="https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-Cameraman_v1">https://huggingface.co/Cseti/LTX2.3-22B_IC-LoRA-Cameraman_v1</a>You can download an example workflow from here: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-cameraman/README.md">https://huggingface.co/datasets/Cseti/ComfyUI-Workflows/blob/main/ltx/2.3/ic-lora-cameraman/README.md</a>As I mentioned in the HF repo, the model has its limitations. Especially with more dynamic movements. In some cases the reference can also turn out to be "too strong", and in those cases it sadly ends up just generating the refence video. I believe these issues are due to the dataset size. If I find the time, I'll try retraining it with a more diverse dataset.

Files with SHA256: 1ac9054d4a085fd4257ed7748c5ce026f05e58c490df93670523fa1f8ad0ae3c